THE EFFECT OF SOLUTE SIZE DISTRIBUTION

ON TH

ROASTING AND LEACHING OF A COMTLEX SULFIDE ORE

by

nChi—shing Hang

Thesis submitted to the Graduate Faculty of the

Virginia Polytechnic Institute

in candidacy for the degree of

MASTER OF SCIECE

in

Mining Engineering

August, 1963

Blacksburg, Virginia

ii

TABLE OF CONTENTS

VPage

I. INTRODUCTION ............................... 1

II. LITERATURE STUDIES ........•................ 8

Occurrence of the Great Gossan Lead

Sulfide Ore ........................... 8

Location .............................. 8

Geology ............................... 8

Mineralogical Characteristics of the

Ore .............................. 9

Ore Reserve ........................... 12

Results of Previous Beneficiation

Investigations ........................ 12

Flotation Investigations .............. 12

Roasting Investigations ............... 13

Leaching Investigations ............... 14

Principles of Roasting Sulfide Ore with a

Fluidized~Bed Kiln .................... 15

General Description ................... 15

Mechanism of Chmical Reaction in Fluidized·

Bed Roasting of Sulfide Ore ...... 16V

Theories of Sulfatizing Roasting of

Sulfide Ore ...................... 17

iiiJ

The Effect of Zinc Ferrite •..„.„...•.. 21

Kinetics and Thermodynamics in a

Fluidized Bed „•..„•„•.„•.•••„.„•.• 23

Formlas for Determination of the

Optimum Conditions in Fluidized·

Bed Roasting of Complex Sulfide

Dres .•.„.......................... 24

Principles of Leaching Copper and Zinc from

the Roasted Ore ...„„•„•„.•.•.......... 27

General Description .„.„..•„••„•„••.••„ 27

Pressure Leaching ...•••••••••••„„••„„. 28

The Relation between Leaching Rate and

Contact Time •...„...„„........•.. 31

Properties of Dilute Solution ..•.••••• 33

Chemical Kinetics ••„.••„••••••••„••••• 35

The Dissolution of Zinc Ferrite in

Dilute Sulfuric Acid ...„..•...••. 38

Elimination of Iron in the Pregnant

Solution with Manganese Dioxide .. 39

Applications of F1uidized—Bed Roasting and

Dilute Sulfuric Acid Leaching ....•...„ 41

Okayama Smelter, Dowa Mining Co. Ltd.,

Japan .•.....••„.•„„...„„••.....„. 41

iv

Kosaka Smelter, Dcwa Mining Co. Ltd.,

Japan ...................•......•. 42

Luilu Metallurgical Plant, Union Miniére

Du Haut·Katanga, Ratanga ......... 44

III. EXPERIMENTAL APPROACH ...........•.....„.... 46

Purposes ....•....•......................... 46

Experimental Procedure ........•......•...„. 47

Principal Experimental Stages ......... 47

Descriptions of the Principal Procedures

and Apparatus ..••................ 47

Experimental Results ..„.•..........„....... 55

Data of Screen Analysis ..............• S5

Roasting Performance .................. 55

Leaching Performance .....•„........... 56

Results of Quantitative Chemical

Analysis ..•..•.•.......••........ 58

IV. STATISTICAL ANALYSIS OF EXPERIMENTAL RESULTS. 81

Description ..•...„•...........•.•..•.•.•... 81

Test for the Particle Size Effects ...„..... 84

Test for the Pulp Density Effects .......... 85

V. DISCUSSION ..................•.....•.....•.. 86

Effect of Particle Size on the Roasting

Process .....•...•.•....•••...........• 86

Particle Size Effect in the Leaching Process. 87

Pulp Density Effect in the Leaching Process.. 91

V

Significance of the Pressure Effect in the

Autoclave Leachiug .................... 93

VI. CONCLUSIONS ................................ 96

VII. SUMMARY .................................... 98

VIII. RECOMENDATIONS FOR FUTURE WORK ............ 102

IX. ACKNOWLEDGEMNTS ........................... 105

X. BIBLIOGRAPHY ............................... 107

XI. VITA ....................................... 111

XII. APPENDICES ................................. 113

vi

LIST OF TABLES

Page

Table 1. Size Distribution of the Roaster Feed..... 60

Table 2. Size Distribution of the Overflow

Ca1cine.............................. 61

Table 3. Size Distribution of the Cyclone Calcine.. 62

Table 4. Size Distribution of the Mixed Ca1cine.... 63

Table 5. Fluidized—Bed Roaating Conditions and

Results.............................. 66

Table 6. Distribution of Calcine................... 67

Table 7. Factorial Arrangement of the Leaching

Tests................................ 68

Table 8. Autoclave Leaching Performance............ 69

Table 9. Chemical Analysis of Ca1cine.............. 70

Table 10. Chemical Analysis of Leach Residue........ 71

Table 11. The Effect of Particle Size and Pulp Density

on the Recovery of Iron in Dilute Sulfuric

Acid Leaching........................ 73

Table 12. The Effect of Particle Size and Pulp Density

on the Extraction Rate of Copper in Dilute

Sulfuric Acid Leaching............... 75

Table 13. The Effect of Particle Size and Pulp Density

on the Extraction Rate of Zinc in Dilute

Sulfuric Acid Leaching............... 77

vii

Table 14. The Effect of Particle Size and Pulp Density

on the Sulfur Content of Lesch Residue . . 79

Table 15. Data of the Factorial Experiment on the

Zinc Extraction Rate .................. 82

Table 16. Analysis of Varisnce Table for Zinc

Extraction Rate ....................... 83

viii

LIST OF FIGURES

Page

Fig. 1 Typical Leaching Curves ..................... 32

Fig. 2 Schematic Diagram showing the Relation between

the Elevation of the Boiling Point,

Lowering of the Freezing Point, and

Lowering of the Vapor Pressure Caused

by the Addition of a Nonvolatile

solute ............................ 34

Fig. 3 Flowsheet of Experimental Procedures ........ 49

Fig. 4 Construction of the Fluidized·Bed Kiln ...... 51

Fig. 5 Construction of the Standard One Gallon

Autoclave ......................... 54

Fig. 6 Cuulative Logarithic Diagram of Screen

Analysis on Roaster Feed and

Calcines .......................... 64

Fig. 7 Direct Logarithic Diagram of Screen Analysis

on Roaster Feed and Calcines ...... 65

Fig. 8 The Effect of Particle Size and Pulp Density

on the Recovery of Iron in Dilute

Sulfuric Acid Leaching ............ 74

Fig. 9 The Effect of Particle Size and Pulp Density

on the Extraction Rate of Copper in

Dilute Sulfuric Acid Leaching ...... 76

ix

Fig. 10 The Effect of Particle Size and Pulp Density

on the Extraction Rate of Zinc in

Dilute Sulfuric Acid Leaching ..... 78

Fig. ll The Effect of Particle Size and Pulp Density

on the Sulfur Content of Leach

Residue ....„.•....••••••„........ 80

x

LIST OF APPENDICES

Page

A. Method for Determination of Moisture and Loss

on Ignition .............•................... 113

B. Method for Determination of Insoluble Matter ..... 114

C. Method for Determination of Sulfur ............... 115

D. Method for Determinatiou of Iron ................. 118

E. Method for Determination of Copper ............... 121

F. Method for Determination of Ziuc ................. 125

G. Example Calculations for the Reocvery of Fe, and

the Extraction Rates of Cu and Zn ........... 130

H. Table of Absolute Vapor Pressure of Water ........ 132

1

I. INTRODUCTION

Since 1951, a series of investigations have been conducted by

the Department of Mining Engineering and the Virginia Engineering

Experiment Station of the Virginia Polytechnic Institute for the

purpose of finding an applicable beneficiation process for a complex

sulfide ore found in Carroll County, Virginia. This deposit is

called the Great Gossan lead. It contains many valuable consti-

tuents including iron, copper, zinc, lead, and sulfur. Favorable

results obtained from past investigations indicated economic poten-

tial for the mineral deposit. If a beneficiation process for the

ore can be developed, the dormant Great Gossan lead will probably

be the largest undeveloped sulfide ore deposit in the Eastern United

States.

Previous beneficiation studies have arrived at the following

conclusions:

A. The significant sulfide minerals are pyrrhotite, marma•

tite, pyrite, and chalcopyrite.

B. Bulk flotation of the sulfide minerals is felt to be the

most feasible concentration process.

C. Differential flotation of the individual sulfldes is prob-

ably not feasible.

D. In order to effect the most complete utilization of the

2

minerals a combined treatment using flotation, roasting and

leaching will probably be most effective.

Other studies on roasting and leaching of the ore with a

f1uidized—bed kiln and an autoclave have indicated:

A. The optimum roasting temperature for the removal of acid

soluble zinc to meet the commercial specification of an

iron ore is between l550°F. and l6S0°F.

B. The optimum roasting temperature for the removal of acid

soluble copper is between l350°F. and l450°F.

C. The effect of leaching time is highly significant, but the

leaching temperature has no appreciable effect.

Many other factors that may affect the roasting and leaching of

the complex sulfide ore are unknown. The effect of the size of solid

particles in the f1uidized—bed kiln and the autoclave, the effect of

pulp density in the autoclave, and the effect of the concentration of

solvent are still to be determined.

This study has been carried out for the purpose of investigating

the size effect of solid particles in the fluidized·bed kiln and the

autoclave. The effect of pulp density in the autoclave has also been

investigated.

Recently the fluidized·bed roasting technique has been utilized

in the preparation of sulfide ores for leaching. Over the past

several years, there have been intensive studies made of leaching

3

in autoclaves where the leaching process can be conducted under high

pressures and high temperatures.

The major objective of most roasting operations is to oxidize

the sulfide psrticles to oxides or sulfates. The oxidation must be

carried out without fusing the charge. Other reasons for roasting

are to volatilize certain impurities and to convert the roaster

charge into porous sinter. Regardless of the type of roaster-

reactor used, the sulfide psrticles must be exposed to a current of

air and ignited. As the air strikes the heated particles they begin

to react as followsz

4 FeS2 + 11 02 ————a> 2 Fe203 + 8 S02 ;AH ¤ • 808,320 Cal...(l)

2 Cu2S + 3 02 ——-a» 2 Cu20 +~2 SO2 ;AH = — 188,900 Ca1...(2)

2 ZnS + 3 02 ————a> 2 ZnO +·2 $02 ;AH = - 222,400 Cal...(3)

Another important chemical reaction that takes place in many

roasting operations is "sulfatizing" or the formation of metallic

sulfates rather than oxides. Hot $102 and Fe2O3, which are almost

always present in roasting operations, act as catalysts to promote

the reaction:

4

Soe of the S02 formed by the burning of sulfur combines with 02 to

form $03. The $03 in turn reacts with metallic oxides to form

sulfates:

Cu0 + $03 ———é> CuS04 ;AH = · 52,300 Cal„..•••.•.„..„.......(6)

Zn0 ;AH = ~ 54,500 Cal.„„.................(7)

The formation of sulfates is promoted by low roasting temperatures

and high S03 concentration„ In general, high roasting temperatures

will decompose sulfates into oxides and $03.

A fluidized-bed kiln·type reactor provides ideal conditions

for gas—solid reaction because each solid particle is in constant

motion and is continually in contact with the mving gas strem•

The calcines obtained as the products of fluidized·bed roast—

ing are introduced to the autoclave and leached with dilute sulfuric

acid. In the autoclave the copper oxide and zinc oxide are dissolved

according to the following:

5

Cu0 + HZSOA-——> CuS04 —&- H20..............................(8)

Zn0 +·H2S04 ————e> ZnS04 + H20..............................(9)

The iron oxide, Fe203, is practically insoluble in the leach

solution and therefore separated as residue. Since zinc ferrite,

ZHFCZÜA, is also insoluble in the dilute sulfuric acid solvent, the

roasting temperature should be low to prevent the formation of zinc

ferrite. Zinc in this form cannot be recovered by leaching, and

its presence prevents the residual iron from meeting comercial

specifications for blast furnace feed• Because the zinc mineral

in the Great Gossan lead 0re is marmatite Zn(Fe,Mn)S, it is very

difficult to prevent the formation of a certain amount of zinc

ferrite.

Chemical kinetics is also involved with this investigation.

Chemical equilibra and thermodynamics involve only the initial

and final states of the chemical reaction. In this investigation,

however, interest lies in the intermediate steps which lead to the

final products. The rates of these intermediate steps and the

factors which determine the rate of reaction, such as concentration,

temperature, pressure, leaching agent and catalysts, are of parti-

cular interest.

6

When the reagents are in two different phases, the reaction

can take place only at the surfaces where the two phases are in

contact• The reaction velocity is then a function of the area of

contact between the phases and the diffusion rates of reagents and

products through the phase boundaries.

It is evident that the surface effect becomes greater as the

degree of subdivision of matter is increased. The characteristics

of fine dissemination of the sulfide minerals in the Great Gossan

Lead Ore have been clarified by previous microscopic studies. Many

of the discrete grains of pyrrhotite, chalcopyrite and marmatite

have been found to be in the size range of 5 microns. It is there-

fore possible that the extent of exposed surface area of the sul-

fide minerals invboth the roasting and leaching processes is a sig-

nificant factor affecting the reactions.

In the process of leaching there are two main reasons for the

high initial rate of dissolution• The concentration of the solvent

is usually at a maximum, and the greatest possible surface of sol-

uble mineral is exposed. As dissolution proceeds, the rate

decreases„ The small minerals dissolve completely, decreasing the

active surface of contact, and the strength of the solvent is de-

pleted„ The minerals that remain may be almost completely encased

7

in gangue minerals th.at present very little surface to the solveut.

These phenomena involve the factors of leach coucentratiou, size

of solid particles and pulp density in the autoclave.

8

II. LITERATUR! STUDIES

Qgcugenee of the Gggg @s_;__sa_g I_Legd gulfidg ggg Qggggit

léssäl-.92The Great Gossau Lead is located in the Blue Ridge Plateau of

Virginia, about 10 to l5 miles uorthwest of the crest of the Blue

Ridge Mountains. The Lead has been traced more or less contimxouelyI

for about 16.5 miles with a ¤n:d.mum width of about 150 feet and an

average width of some 20 feet. The main Lead has its southern

terminus near Galax in western Carroll County and trends n0rtheast·

ward through Carroll County to a point near Sylvatus, a few miles

west of the Floyd County line. Offset: of the main Lead are found

in the Sutherland and Toutray Leads of Floyd County, and a major

offeet is in the exttuue eastern part of Grsyoon County (1-*2)-

mlmIn general, the Great Gossan Lead is oriented with the strike

N.25·60°R. and dip 45°S•E• The mein Lead and offset lie in a com-

plex of crystalline schists and gneisses (1). The schists and

gneisses are intruded by umsses of relatively unmetamorphosed dio-

rita, and by small bodies of granite and syen:I.ta• A fine-grained

mica and sericite schist ls the main country rock throughout the

main Lead region. The component mineral grains of the schist are

small, although they are usually in aggregates of a size coarser

than that of the sulfide grains that are sometimes found as discrete

9

grains in the schistose folds. The major structure responsible for

the emplacement of the ore-bearing fluids along the main Lead may

have several, or a combination of several origins. Detailed field

work will be necessary to work out the structural pattern. After

the formation of the structural framework, the mineralizing fluids

ascended along the major fractures and began the replacement and

alteration of the country rock iumdiately adjacent to the rock

opeuings, forming a hypothermal deposit of primary aulfides. Follow-

ing the emplacement of the primary sulfides at depth, erosion with

subsequent weathering and leaching of the upper part of the vein

zone formed a gossan with its usual oxides and carbonate ores of

copper and iron. Secondary enrichment of copper sulfides occurred

immediately below the water table. Although C. S. Ross (3) indi-

cated a late Paleozoic age for the formation of the Great Gossan Lead,

the age of the deposit is still unknown.

Mineralogical Characteristics of the Ore

A. Component Minerale. The primary eulfides of economic

eignificance are ae follows (in order of decreaaing

abundance):

l. Pyrrhotite - Fe8S9

2. Marmatite · Zn(Fe,Mh)S

3. Pyrite — FeS2

4. Chalcopyrite·

CuFeS2

10

ln addition to these, galena (PbS), arsenopyrite (FeAsS), 1ö1l1n~

gite (FeAs2), and/or leucopyrite (Fe3As4), tetrahedrite (essentially

a copper antimny sulfide) and/or tennanite (similar to tetrahedrite

with arsenic replacing the antimony), cubanite (CuFe2S3), and iron

oxides (magnetite - Fe30A, hematite · Fe203, limonite • Fe203.nH20)

have been identified. The letter minerals are generally present to

the extent of less than one percent. The h1gh•grade secondary minerals

which were mined early in the history of the lead and of which only

minor quantities may still be seen are chalcocite and covellite.

Earlier the presenee of oxidized copper ores, including carbonates,

had been reported in the gossan. The significant gangue minerals

are quartz, carbonates, chlorite, micas and garnets.

B. Chemical Properties. The relative abundance of pyrrhotlte

in the sulfides varies from 70 to 951. An analysis of a nearly pure

specimen of pyrrhotite gives its composition as Fe 60.81 and S 39.21.

The z1nc•bear1ng sulfide mineral has been determined to be marmatite

rather than sphalerite. An analysis of a minus 325•mesh flotation

concentrate of the marmatite shows the average composition: Zn,

55.331; Fe, 8.291; Cu, 1.241; Mn, 0.201; and S, 35.051. Marmatite

accounts for 3 to 81 of the sulfides. Pyrite, representing 1 to 201

of the sulfide mlnerals, is noted in the mere heavily mineralized

sections where it replaces an equivalent quantity of pyrrhotlte.

The chalcopyrite distribution in the sulfide is fairly uniform and

averages 2 to 31. An analysis of the Great Gossan lead head sample

11

show its composition es follows (4):

Element Weight L

Fe 35.52

S 25.37

Zn 1.96

Cu 0.63

Pb 0.QQTotal 63.82

Gangue (by difference)36.18

C. Physical Properties. Previous qualitative microscopic

studies on the ore have reveeled the peculiar physical features of

the sulfide minerals present (1). The sulfide-gangue intergrowths

feature smooth boundaries of the minerals. The grain sizes for the

sulfide and gangue aggregates average about 65-esh and 28·mesh

respectively. Notable exceptions are the garnets and silicates

which have hair-line veinlets of aulfides - principally chalcopyrite.

Many of the angue graina show 5-10 micron inclusions of sulfides-

principally marmatite. The boundary contacts between the various

sulfides very from smooth to interlaced. Pyrrhotite and pyrite occur

in discrete grains as coarse as 35-mesh, although the marmatite con-

tains pyrrhotite inclusions of 5 microns or less. Discrete marmatite

grains are usually smaller than 150-mesh. Microscopic examination

of the typical marmatite—pyrrhotite relationship has indicated that

about 50 per cent of the zinc sulfide grains will be found in the

size range of 5 to 15 microns. The chalcopyrite is usually finer

12

than the marmstite. Discrete chalcopyrite particles average less

than 10 microns. Galeua occurs largely as very small locked particles

of 5-10 microns in the marmatite. The: iron oxides and arsenopyrite

generally appear as 5-15 micron discrete grains in the gangue. The

tetrahedrite-·tennanite occurs with a grain size of 10 microns or

less, and always as incluaions in galeua.

äeäeumesIt has been estimsted that the Great Gossau Lead might contain

over 200 million tons of ore reserves (4). Because of this trennen-

dous size, it has the potential to be a major source of iron ore.

The main source of iron would be from the pyrrhotite, which is the

principal component mineral of the ore deposit. Also, coustderable

quantities of eopper, zinc, lead, and sulfur could be obtained from

the minor component minerals • pyrite, chalcopyrite, marmatite, and

galena.

[logtion ggggtiggtgog

A series of flotation studies, including bulk sulfide flotatiou,

pilot plant tests, and differential flotation, have been conducted

at the Virginia Pplytechnic Institute. Statistical analyses of the

experimental results have shown:

A. The Lead ore is emenable to a relatively simple bulk sul-

fide flotation procedure. Theoretically, bulk sulfide concentrates ,

13

recovering 98% of the sulfides and assaying 53.5% iron and 38.5%

sulfur, can be obtained. For a recovery comarable to the theoretical

value, a bulk sulfide concentrate from the Betty Baker area, represen—

tative of the most complex material found in the deposit, will assay

49.5% iron, 36.0% sulfur, 3.3% zinc, 0.78% copper, and 0.67% lead(1).

B. Differential flotation of the individual sulfide requires

extremely fine grinding, which may be uneconomical. The sulfides

must be ground finer than minus 325-mesh in order to liberate the

marmatite, chalcopyrite, and galena from the iron sulfide. Differen-

tial flotation from a reground bulk sulfide concentrate does not offer

much promise. Differential flotation after depressing marmatite by

using sodium sulfide (Na2SO3) offers the best promise of making high-

grsde iron, copper-lead, and zinc concentrates (1).

C. For overall high recoveries a treatment combining roast-

ing and leaching of a bulk sulfide concentrate is believed to be

most promising.

Roasting Investiggtions

Studies with applications of roasting and leaching techniques

have been carried out at the Virginia Polytechnic Institute. Explora-

tory roasting tests were first performed in a 1-inch diameter fluidized·

bed column. The preliminary test results justified continuation of

the work with application of a 4-inch diameter fluidized-bed kiln.

Studies on the roasting temperature effect in fluidized-bed kilns

were conducted by C. H. Shiuh(5). The roasting conditions used by

14

Shiuh in his experimets were: feed rate, 8 grams per minute; air

supply, l cu. ft. per mnute; range of roasting temperature, lO50°F.

to 1650°F. A mixture of bulk sulfide flotation concentrates and

table concentrates was used for roasting. An optimum range of

roasting temperature for the reoval of zinc content from roasted

calcine by a subsequent leaching process was determined to be 1550-

l650°F. An optimum range of roasting temperature for the removal of

copper content from roasted calcine by a subsequent leaching process

was determined as 1350·l450°F.

geaching Investiggtions

Earlier tests of leaching roasted calcine under static condi-

tions in uniquely designed pressure vessels were conducted by W. E.

Foreman, staff member of the Department of Mining Engineering,

Virginia Polytechnic Institute (4). A statistical analysis of the

test results indicated that the significance of the errors exceeded

that of the variables and no clear cut conclusion was possible. Con-

sequently, leaching experiments, utilizing a one-gallon stirred auto-

clave, have been carried out. In connection with the f1uidized•bed

roasting experiments, Shiuh has atudied the effect of leaching time

and temperature in the process of leaching the fluidized·bed calcine‘

with dilute sulfuric acid solvent in the autoclave (5). The leaching

conditions used in the experiments were: solvent, 13 ml. concentrated

sulfuric acid diluted with 2000 ml. distilled water; pulp density,

200 grams of roasted calcine per 2000 ml. distilled water plus 13 ml.

sulfuric acid; leaching temerature, 250°F. and 450°F; leaching time,

15

1-hour single leaching and 1-hour double leaching (totalling 2 hours

for each double leaching). The experimental results indicated that

the effects of leaching time and roasting temperature were signifi-

cant, but the leaching temperature had no appreciable effect.

Fluidiged-Bed Kiln

Gengrgl Qgscripgign

Sulfide ore or cuncentrate is not usually in a form to undergo

direct reduction and thus conversion to another state is necessary.

This is commonly affected by roasting, which may be defined as heat-

ing below the fusion point with an access of air. Although many

other avenues of approach were explored, oxidizing and/or sulfatiz-

ing roasting followed by leaching with dilute sulfuric acid solution

or water appears to hold promise as a successful procedure for treat-

ing cupriferous and nickeliferous pyrrhotites. The method of fluidi-

zation or fluosolids roasting, in which finely divided solid materials

are kept in suspension by a rising current of air, was developed

by the Dorr Company in 1945. This process is now in use for roast-

ing zinc concentrates, arsenopyrite-gold concentrates and for

production of sulfur dioxide from iron pyrite. Copper sulfides,

especially those containing large amounts of iron sulfides, pyrite

and pyrrhotite, can be given a sulfatizing roast which makes the

copper soluble in water or weak acid (6’7). Close temperature-

control euables a 90 per cent water-soluble product to be obtained

with only 1.5% iron going into solution. This opens up interesting

16

poesibilities in that copper sulfide concentrates could be roasted,

leached, and treated elctrolytically, by—passing the smelter

altogether•

ygggggism of Qhgglcgl ggggtions in Fluidized-Bed Roagting of Sulfide

ggg.

The fluidized•bed kiln-type reactor provides an ideal condi-

tion for gas-solid reaction beause each solid particle is in

constant motion and is continually in contact with the moving gas

stream on all sides. As previously shown by the aquations (1)

through (7), the iron sulfide, copper sulfide, and zic sulfidel

are converted into iron oxide, copper oxide, and zinc oxide

respectively by the oxidation reactions, and some of the oxides

are converted into iron sulfate, copper sulfate, and zinc sulfate

by the sulfatizing reactions during the process of fluidization

roasting„ As the reagents are in two different phases- gasous

and solid, the reaction can take place only at the surfaces where

the two phases are in contact. The reaction velocity is then a

function of the area of contact between the phases and the diffusion

rates of reagents and products through thc phase boundaries. Since

these oxidation and sulfatizlng reactions give off a large amount

of reaction heat, the roasting process is self·sustaining once

the reactions begin, provide a proper temperature control is

maintaind. In many cases higher temperstures would cause fusion,

volatilisation, or soe other phase change that must be avoided in

17

a particular operation. Reaction: such as equations (2) and (3),

when conducted in meeting furnaces, wat be maintained at a low

tanperature level so that none of the solid material fuaes. Regu-

lation of temperature often provides s measure of control over

other reactions involving gangue minerals or impurities. The

temperature may also effect the nature of the primary reaction.

When sulfides are roasted in air, they usually form metal oxides

and 802 as shown by the equations (1), (2) and (3). At low tempera-

tures, however, the oxygen may combine directly with the sulfide

to form a sulfate. Some authorities feel that in all cases the

sulfates form first and that the high txerature then decomposes

( the sulfates into $02 and metal oxidesg in any event the roasting

temperature is very important in detemining the nature of the com-

pound formed. It is possible to attain selective roasting of

mixtures of different aulfides by proper tmperatura control (8).

Theorig of gglfggizigg Roastgg of gulfidg Qge

Some authorities have claimed different theories related to

the formation of sulfates in the process of sulfatizing roasting of

sulfide ores. Ashcroft (9*10), working with Rhodesian ores, found

that the reaction in sulfatizing roasting proceeds primarily and

definitely to the formation of copper sulfatea and not oxides, and

that the formation of oxidea and liberation of 802 are secondary

reactions caused by the decomposition of previwsly formed sulfates.

He sxmmarized bis results as:

18

A. Iron oxide probably acts as an efficient catalyzer in

the formation of copper, zinc, and nickel sulfates.

B. At temperatures above 550°C. or thereabout, formation of

cupric ferrite (Cu0·Fe203) begins when copper oxide and iron oxide

are brought into juxtaposition, end at 700°C. the reaction is

prohibitively rapid. Cupric ferrite is insoluble in any dilute

acid·leaching normally used in copper leaching.

C. The ordinary rabbled furnaca is inimical to the production

of all water•soluble copper sulfate by reason of the omnipresent

reaction:

CuS04 + Cus + 02-*-2CuO + 2802 ..........................(l0)

D. A reaction similar to the above resulting in the decompo-

sition of sulfates is apt to occur in the presence of carbon or

of any reducing gas, such as flue gas:

CuSO4 + CO———*· Cu0 + S02 + CO2 .........................(ll)

E. Iron sulfate is not, as is usually atated, decomposed

at a tmperature so far below the decomposition temperature of

copper sulfate that a mere roasting at a carefully regulated

temperature may be employed to produce, quantitatively, soluble

copper and insoluble iron.

Peretti (11) refuted the claim of Ashcroft that copper oxide

production resulted only from secondary decomosition of sulfates

that were formed as primary products. He showed that a layer of

Cu20 appeared directly adjacent to the Cu2S during roasting of

19

briquettes of cupric sulfide, CuS. The reactions proposed were:

2CuS —-4> Cu2S +·% S2 .......................„............(12)

-;-52 + O2 —> SO2 •••••-.••••••••••••••••••••••••••••••••(13)

Cu2S +-3-02 ——> Cu20 + 802 ..........•..............„.„(l4)

01120 +·é 02 ——> 2CuO ...................................(15)

At temperatures above 663°C., Cu0 was the only final solid phase

reported. Below 663°C., increasing amounts of sulfate were found

mixed with the Cu0. When the outer layer was wholly Cu0, it was

black, porous, and easy to cut with a knife. The specimen contain-

ing sulfate became harder and less porous as the sulfate content

increased until at 478°C. and 430°C., the pieces became so hard and

dense as the holding time increased that the reaction practically

ceased in a Cu0 layer between 3 and 4 millimeters in thickness.

Investigation by McCabe and Morgan(12) supports the claim of

Peretti. They reported the following sequence of products, beginn-

ing at the sulfide surface, when roasting discs of synthetic chal-

cocite at 600°C.: Cu20, a mixture of Cu20 and CuSO4, CuS04,

Cu0•CuS04, and CuO at the outer surface.

Thornhill and Pidgeon (13) studied the roasting of sulfide

and reported the formation of firm, protective oxide shells when

roasting pentlandite particles at temperatures below 650°C. They

suggested that much of the oxide growth took place by ionic migra-

tion through the continuous oxide layer. The oxidation rate was

found to increase with fineness of grind when roasting pyrrhotite.

Whereas 65 to 80•mesh particles roasted linearly at 550°C., the

20

oxidation rate of 30 to 40—mesh particles was retarded after the

formation of about 160 microns thickness of oxide. The roasted

pyrrhotite particles were deeply penetrated by oxide column.

Wadsworth's study (14) of sulfatizing Cu2S indicated that

oxidation reactions generally take place by diffusion. However,

entrapped gases in certain temperature ranges almost stop oxidation

before the process is completed.

Lewis (15) determined that the optimu temperature for oxida·

tion of chalcocite to give water-soluble copper ia about 450°C.

with both air and oxygen and that oxygen is a more effective oxidi-

zing agent than air for conversion to water-soluble substances. At

about 450°C. the following reactions were confirmed:

2Cu20 + Cu2S -———ä• 6Cu + $02 ............................(l6)

2CuO + Cu2S 4Cu + 802 .............................(l7)

Fursman (16) also determined that basic copper sulfate begins

to form at a little above 300°C. and in the presence of cupric

oxide the temerature of formation is probably mch lower because

of the reaction:

Cu0 + CuS04 ————+> Cu0·CuS04 .............................(18)

Observations of Baroch (17), working with copper ore at Bagdad,

Arizona,indicated that reactions take place in three distinct stages,

which may overlap to some extent, but which take place only in the

U order given. Each stage requires a definite temperature range and

the best oxidation conditions possible in order to be completed with-

in practical time limits. He was successful in almost comletely

2l

transforming the copper into a water-soluble state as by:

A. First stage, 370° to 380°C., decomposed all but a trace

of true sulfides while the sulfur was driven off or oxidized to

form sulfates. Temperatures above 455°C. should be avoided before

this stage is ccpleted, otherwise a true magnetite will be formed

that is difficult to reoxidize tc the less soluble ferric form. The

probable reaction is:

Fes + l0Fe203 SO2 ......................(l9)

B. Second stage, 455° to S40°C., is the oxidizing or demag-

netizing period during which the dcsired higher oxides of copper

and iron are formed.

C. Third stage, S90°C., selectively destroyed the water-soluble

sulfates of iron.

The Effect of Zine Ferrite

Zinc sulfide will not dissolve in the leaching solution nor

will it reduce in the zinc retort. When roasting zinc concentrates

that contain iron there is always a certain amount of zinc ferrite,

Zn0•Fe203, formd. The zinc ferrite is as undesirable as the sul-

fide because it will not dissolve in the leach solution nor reduce

in the retort. The roasting temperature for zinc concentrate is low

in order to form soe zinc sulfate (small amounts of sulfate are

desirable in feed for zinc leaching) and also to prevent the form-

tion of zinc ferrite. However, when the zinc mineral is marmatite,

(Zn,Fe)S, it is very difficult to prevent the formation of a certain

22

amount of zinc ferrite. One of the reasons for the trend toward

fluidized-bed roasting is that it may minimize the formation of

ferrite.

Nii, Kimura, and Hisamatsu (18) have studied the formation of

zinc ferrite from a stoichiometric mixture of ferric oxide and

zinc oxide. Their iuvestigations indicated that the ferric oxide,

which was ignited at higher temperatures, showed lower reactivity

with zinc oxide, and the rate of formation of ferrite decreased

with the increase in ignition temperature. The effect of ignition

temperature was more apparnt when the ferrite ie formad at low

temperature. The activation nergiea of formation of ferrite

increase from 39 to 62 kcal./mol. gradually, according to the

increase in ignition temperature.

Latest report revealed that trials had been successfully

conductad at the hydrometallurgical plant of Cerro de Paaco Corp.

in La Oroya, Peru, to roast marmatite concetrates at a tempera-

ture as high as 1l50°C. in a fluidized bed. At this high tempera-

ture, the insoluble zin ferrite, formed from the Zn and Fe203

in the roaster fee at tamperatures above 650°C., was decomosed

into acid soluble Znß and insoluble Fe304. The extraction of zinc

by leaching the calcines roasted at l1S0°C. idicated 2-3% more than

that by leaching the calcines roasted at 9S0°C.(19).

23

Kggggics ggd lhegmodyggmicg ig g Fluggized Bed

The motion of solid bodies in a fluidized bed has been studied

by Yu. P. Gupalo(20)

with a radiometric method. The experimental

results indicate that in the presence of an ascending stream near

the column axis, the medium in a fluidized bed can be regarded as

a ewtonian fluid in which the viscosity does not change with the

time rate of deformation to a first approximation.

It is known that the size of the particles and their specific

gravity influence heat exchange between a gas and particles in a

fluidized bed. The heat-transfer coefficient of the particles

toward the gas, other conditions being equal, increases with an

increase in the size of the particles and their specific gravity.

Ge1‘perin and Ainshtein (21) explained this phenomenon, theoreti-

cally and experimentally, and proposed the following relationships

on the influence of the size and specific gravity of solid particles

on their heat transfer coefficient toward a gas in a fluidized bed:

A. The coefficient of heat transfer from particles of a

given size and specific gravity to a gas is a function of the

velocity of the gas and is expressed by the equation:

3 == han ..•.•..••.....••..•..••.••••...•...••••••.•.•.••.....(20)

where, a • the heat-transfer coefficient of the particles toward

the gas.

A -constant.

w • mean velocity of the gas.

n - constat.

24

B. If the "effective" coefficient of heat transfer (a) is

determined, the total quantity of heat transmitted from the gas to

the particles can be expressed by the equation:

q = a•Ax•f·N ............................................(21)

where,M

q - the total quantity of heat transmitted from the gas to the

particles.

At - the difference in temperatures between the gas and the—

surface of the particles.

f -the surface of one particle.

N • the total quantity of particles.

C. The dependence of the heat—transfer coefficient on the

size of the particles and their specific gravity is expressed by

the power function:

s A, apxq ..............................................(22)where,

d · the size of the particles.

r — the specific gravity of the particles.

and,

p/q = 3.

Formulas for Determination of the Optium Conditions in F1uidized~

Bed Roasting of Cggplex Sulfide Ores

In general, the major purpose of the fluidized·bed roasting of

complex sulfide ore is to extract copper, zinc and other valuable

25

components from the calcine. High extraction of metallic comonents

can be attained by the precise control of roasting conditions.

Nakajima, Nishihara and Kondo(22),

in a work on the application

of fluidized-bed roasting for hydrometallurgy, propcsed a series

of formulas to determine the optimum conditions in the sulfatizing

roasting of pyrrhotite, chalcopyrite, and other sulfide minerals and

metallurgical wastes, suitable for the extraction of copper and

zinc. The proposed formulas are shown below.

A. Determination of air supply:

A = géäéäl B(1 - a) +- { k(b + a/2) - B }· ($02)] ..(23)

where,

A · quantity of air supply per 1 kg. dry ore, in Nmg.

(S) · sulfur content in l kg. dry ore, in kg.

($02)-

voluetric ratio of $02 content in the roasting gas.

a - conversion rate of $03 formed from S02.

B - extraction of sulfur.

k · air consumption rate

k S air consggption at sulfur extraction = Bair consuption at complete extractionof sulfur by oxidation

b • oxygen factor, that is the number of moles of oxygen

required to form l mole of SO2.

26

B. Determination of feed rate:

F „ fg U Q6 X

273ATdL=--6-—_ OICOIOOOIOOIOIOOIOOIOOOOIOOOIOOOOl•IIlIOI•IIOlO(25)

__ V60;)I: kgb + agz} - 16“ 0.7B(1 - aus 1 na - 6;

‘S°2)•··•····(26)

where,

AT — cross-section area of the fluidized-bed kiln, in mz.

Uf - fluidizing velocity of the ore particles, in cm/sec.

t - roasting temperature, °C.

0 - average retention time of the ore in the kiln, hr.

d - apparent specific gravity of the ore, in kg/m3.

L • thickness of the filling layer in static state, m.

V · air supply per unit time, in Nmg/hr.

C. Determination of the concentrations of gases in the kiln:

The maximum amount of copper, zinc and the valuable comonents from

the calcine will be extracted when the concentrations of gases in

the kiln meet the following:22 3: [ :3 ]CD OII•I|OOI§OIlOIO•IO•(27)

where,”

kC b(1 - q) + q

¤·=·1i(b+—l)(1- q)+<1B 2

q - concentration of oxygen in the intake gas, for air q = 0.21.

K - the equilibrium constant for the reaction, SO2 +·% 02:: $03.

27

Principles of Leaching Copper and Zinc From Roasted Ore

General Description

Hydrometallurgical methods have made considerable progress in

recent years because of the improveents in practice and because

it is particularly suited to the treatment of many low-grade ores.

Gold ore, oxidized uranium and vanadium ores, and oxidized copper

ores are all treated by hydrometallurgy. In addition hydro·

metallurgy is a very important method of treating sulfide zinc

concentrates. Recent studies on the application of hydrometallurgy

to the preparation of pyrrhotite as an iron ore have also indicated

great promise. ggpching or lixiviation, which dissolve the metal,

and pgecipitation, whereby the dissolved metal is reoved from the

. leaching solution, are the two essential steps in a hydrometallurgical

process. Calcine, an oxidized material formed by roasting sulfide

concentrates, is one of the materials suitable for leaching. Many

different solvents have been used for leaching purposes. Some have

been very successful while others have not. In general the solvent

to be used for leaching must possess the following characteristics:

(1) it must dissolve the ore minerals rapdily enough to make com-

- mercial extraction possible, and it should not attack the gangue

minerals, (ii) it must be cheap and readily obtainable in large

quantitiea, and (iii) if possible, it should be such that it

can be regenerated in the subsequent precipitation prbcess. Water

can be used for leaching such minerals as CuS04 and ZnS04. Dilute

28

sulfuric acid is used in the dissolution of oxidized copper and

zinc according to the reactions shown previously by the equationa

(8) and (9). An aqueous solution of sulfuric acid is commonly

used as the leaching agent in the treatment of calcine produced by

roasting sulfide ore.

Pressure gggching

For the past several years there has been an intensive effort

made to leach in autoclaves under high pressures and with higher

temperaturés. High pressures are advantageous when gaseous reagents

such as oxygen and hydrogen are used because the amount of gas kept

in solution depends upon the pressure of the overlying gas. Recent

work has shown that it is possible to dissolve sulfides directly

in water, acid, or ammoniacal solutions if they are heated sufficiently

and maintained under high enough pressure so that sufficient oxygen

is forced into solution. The previous leaching methods for sulfides

required roasting to convert them to soluble compounds (23).

In a study concerning oxygen pressure leaching of complex sul-

fide concentrate utilizing an autoclave, Umetsu, Tozawa and

Sasaki (24) reported their experimental results as follows:

A. An aqueous pulp of the complex sulfide concentrates (-200-

mesh) with a composition of 10.6% Cu, 20.5% Fe and 34.7% S was

treated with oxygen at elevated temperature and pressures in a

liter stainless steel stirred autoclave.

29

B. Ninety-nine per cent of the zinc and 95% of the copper

in the concentrates were extracted under the following conditions:

1. pulp density: 10.7% solid.

2. leaching time: 1 hour.

3. temperature: 185°C.

4. total pressure: 40 atm.

C. Ninety-nine per cent of the zinc and 90% of the copper

were leached at l70°C., under 20 atm. in 4 hours.

D. At a temperature of l85°C. and a total pressure of 40

atm., pulp density and leaching time were varied as:

1. pulp density: 10.7-24.4% solid (30-80 grams in

250 c.c. water).

2. leaching time: 1-5 hours.

The results indicated that more than 95% of the eine in the concen-

trates were leached under these conditions, and 90% of the copper

was extracted when the pulp density was 24.2% solid with a reaction

time of 4 hours.

E. The amount of iron leached increased with the increase in

pulp density. However, the concentration of iron in the pregnant

solution tended to be a constant value for every pulp density when

the leaching time became longer.

Pressure leaching has two advantages over leaching at

atmospheric pressure. It permits much greater concentration of

gaseous reagents such as 02 and H2, and it permits higher operating

temperatures which usually means faster dissolution.

30

In some cases both of these factors are important, while in

other cases only one of them may be significant. It is of interest

to note some pertinent facts about the solubility of substances in

superheated water. The critical point for water is 374°C. and 218

atmospheres (about 3200 psi). At this pressure and temperature

liquid water and steam becomes a single phase. At temperatures between

the freezing point of water and its normal boiling point the solubi-

lity of the permanent gases decreases as the temperature rises, so

that it is possible to practically expel all the dissolved air from

water by boiling it. On the other hand, many salts and bases show

increasing solubility when the temperature rises. At the higher

temperatures that can be maintained in an autoclave, these effects

are reversed. As water approaches the critical point, the solubility

of salts approaches zero and the solubility of gas becomes practi-

cally infinite. Presumably this is because all gases are mutually

soluble,amd as liquid water approaches the critical point, it

becomes more like a gas. Perhaps for the same reason it loses its

solvent power for salts. This means that leaching and precipitation

cannot be carried on at excessively high temperatures because the

materials will not stay in solution and because excessively high

temperatures require extremely high pressures. It is therefore

conceivable that the chemical reactions that take place in pressure

leaching are often quite complex.

31

The Relation Between Leaching Rate and Contact Time

The curves shown in Fig. l are typical of the curves obtained

when the laboratory data of extraction rate against leaching time

are plotted. The vertical coordinate is the per cent of the metal

dissolved and the horizontal coordinate is time. Such curves have

a characteristic shape • they rise rather rapidly from zero time

and then flatten off and approach the 100 per cent asymptotically.

In some cases the asymptote may be below 100 per cent if some of

the metal is in an insoluble form or completely encased in gangue

particles. When leaching is speeded up by using higher temperatures

or greater pressures, the curves rise more rapidly and approach the

limit more quickly, but they still have the same characteristic

shape. There are two main reasons for the high initial rate of

dissolution; the concentration of the solvent is usually at a

maximum, and the greatest possible surface of soluble mineral is

exposed. As dissolution proceeds the rate slows down because the

sall minerals dissolve cempletely and thus greatly decreases the

active surface of contact, the strength of the solvent is depleted,

and the minerals that remain may be almost completely encased in

gangue minerals so that they present very little surface to the

solvent. Mat of the dissolution takes place early. In practice

100 per cent extraction cannot be achieved. It is necessary to

determine the extraction time that will give the maximum economic

return.

32

$100

g 80

ä60

= Z';E v

0

LEACHIII} '1'III ——>

HG. 1• '1'YPICAL LELCHING CUBYES

33

Qgggerties of Dilgte Solgtiggs

In this investigation, dilute sulfuric acid was used to leach

copper and zinc contents from the roasted ore in an autoclave. It

is therefore of interest to review the properties of dilute

solutions.

There are four properties generally applicable to dilute

solutions (25):

A. Iowering of the vapor pressure.

B. Elevation of the boiling point.

C. Depression of the freezing point.

D. Osmotic pressure.

The first two properties are considered for the solutions of non-

volatile solutes while the last two properties are considered for

solutions in general. However, these four properties are closely

related. Fig. 2 is a schematic diagram showing the relation between

the elevation of the boiling point, lowering of the freezing point,

and lowering of the vapor pressure caused by the addition of a non-

volatile solute. It is noted that when a solution is separated

from the solvent by a membrane which is permeable to the solveut but

not to the solute, the solvent flows through the membrane into the

solution where the chemical potential of the solvent is lower.

This process is known as osmosis. An excess of pressure must be

34

IUSION OURVI VAPORIZATION CURVEFOR SOLUTION FOR PURE SOLVENT

/* "° E *' * '* E E * " *' "' E

II I msxos emma /IIFOR_II I /PURE SOLVENT // I I

5 II I / I I1 I I I I / \I_I\VAPORIZATION

I I I / CURVE FORI I I / I I SOLUTIONI I I // I II I I / I I

°° I I I // I II I I OF I I

I I I // I ISUBLINATIOH I ·I’ I IÖURVE NR I I I

PURE SOLVENT I I I II I BOILING POINT I I BOILING POINT

OOF SOLVENT \ / OF SOLUTION

/7 %FRIEZING POINT I I FREEZING POINT ÜbOF SOLUTION AT! OF SOLVEIT A

TENPERATURE

FIG• 2• SCHEIIATIC DIAGRAN BHOWING TE RELATION BETIEEN THEELEVATION OF TE BOILINO POINT, LOUERINU OF TE FEEEZIM POINT,

AND LONERING OF TE VAPOR PRESSURE CAUSED BY TE ADDITION OF ANONVOLATILE SOLUTE• TE DIAORAK IS NOT TO SOALE•

35appliedto the solution to prevent this flow of solvent through the

membrane. The osmotic pressure is the pressure difference across

the membrane required to prevent spontaneous flow in either direc-

tion across the membrane.

At equilibrium the chemical potential of pure solvent at l atm.

pressure (ul) is equal to the chemical potential of solvent in the

compressed solution (u2):

ul = u2 ..................................................(28)

The chemical potential of the solvent in the solution is less than

that of the pure solvent at the same pressure and temperature, but

the chemical potential of the solvent in the solution can be raised

by applying pressure. When the applied pressure is equal to the

osmotic pressure, the two effects exactly counterbalance one

another, so that the equation (28) is followed.

Chemical Kinetics

Chemical kinetics is greatly involved in this investigation.

Chemical equilibria and thermodynamics involve only the initial and

final states of the chemical reaction. However, we are interested

in the intermediate steps which lead to final products.

The rate of a chemical reaction generally depends upon the

concentrations of the reactants. The rate may also depend upon the

concentrations of other substances not involved in the stoichio-

metric equation.

36

The rate at which solide dissolve in an unsaturated solution

depends upon the concentration gradient in the immediate vicinity

of the solid and upon the surface area of the solid. The latter

fact is sometimes called Wenzel'e law(26).

The first substantial

treatment of the problem of reaction velocity at a solid-liquid

interface, by Noyes and Whitney, led to the following relation:

gi = m(cB - c) „.•.....•.•......•......•...•...•..•......(29)

in which c is the concentrstion in the bulk of the solution at time

t, cs is the eaturation concentration, and m is a constant of pro-

porti¤ality• The constant requires that the experimental condi-

tions remain the same, specifically that the surface area of the

solid and the thickness of the viscously flowing film adjacent to '

the solid do not change. It is convenient to express the change

in concentration with time, dc/dt, in terms of the velocity constantA

(K) and the ratio of the volume of solution to the exposed inter-

face (-g) .

gf " 7,%- (CS - C) ..................................(so)Rearrangement of variables and integration gives:

1< =§§ ln(61)

where,

K - the velocity constant, in centimeters per minute.

V · the volume of solution, in cubic centimeters.

S — the exposed interface, in square centimeters.

37

9 - the elapsed time, in minutes.

Cs — the saturation concentration, in any units as grams per

liter.

cl - the initial concentration, in any units as grams per liter.

cz - the terminal concentration, in any units as grams per

liter.

In a binary system consisting of solid and liquid phases, the

rate of solution of solute in a solvent is greatly accelerated by

pulverizing the material to expose a larger area to the solvent.

Quite apart from this effect, the actual equilibrium solubility of

very small crystals is greater than that of large crystals. The

greater solubility of the very small crystals is analogous to the

higher vapor pressure of small droplets of liquid. This increase

in solubility with decreasing size is not a significant factor

unless the crystals are extremely small (27).

Ore to be leached must be crushed and ground to such a size

that the solvent can readily act on the soluble minerals. Grinding

is usually unnecessary for porous qxidized ores.

Ior leaching purposes ore products are divided into two classes -

slimes and sands. Slimcs are ore products that are so finely ground

that they tend to pack into a vat or tank and prevet the free

circulation of liquid through the interstices of the bed of ore.

Sands are products that are coarse enough to permit circulation of

the solvent through the void spaces between the ore particles. These

38

two materials require different methods of attack. Slime leaching

must be carried out by agitating the slime and solvent together in

a suitable agitating tank. Sand leaching is done by alloing the

solvent to percolate through the ore to be treated.

The Qissolution of Zinc gerrite in Dilute Sulfuric Acid

As described previously, zinc ferrite is an undesirable material

in the leaching process because it is practically insoluble in dilute

sulfuric acid solution. In connection with the study on the forma-

tion of zinc ferrite, Nii and his associates (18)have investiated

the dissolution of zinc ferrite in 9% dilute sulfuric acid under a

temperature of 50 to 80°C. The maximum time of dissolution has been

carried on as long as 1,215 minutes. The zinc ferrite, used in the

tests, was formed by igniting a mixture of ferric oxide and zinc

oxide. Their observations indicated that the rate of dissolution

of zins ferrite in dilute sulfuric acid was lowered by raising the

preignition temperature of ferric oxide and the formation tempera-

ture of zinc ferrite. The dissolution rate was proportional to the

surface area of ferrite particles, and the activation energies of

dissolution were 15 to 18 kcal./ol.

It should be emphasized that one of the ultimate objectives of

this investigation is to prepare the Great Gossan Lead complex sul-

fide ore as a commercial iron ore suitable for blast furnace feed.

Zinc is a critical impurity since it will attack the refractory

linings in the furnace with strong corrosive effects, and the fumes

39

formed by the vaporization of zinc will precipitate in the cooler

sections of the furnace. Therefore elimination of the formation

of zinc ferrite in the roasting process is important for obtaining

a higher extractiou of zinc in the subsequent leaching operation.

Elimiggtion of Iron in the Pregggnt Solution with Manggnese Dioxide

Prevention or eliminstion of the dissolution of iron in the

leaching solution is necessary in order to:

A. Maximize the recovery of iron when the leached residues

are to be used as iron ore.

B. Eliminate the impurities in the electrolyte when electro-

lysis is used to precipitate copper, zinc and other vsluable

metals from the leach solution.

At the Kosaka Smelter, Dowa Mining Co. Ltd., Japan, manganese

dioxide powder has been used to eliminate ferrous sulfate in the

pregnant solution in the process of treating Cu—Zn concentrates

with sulfatizing roasting, dilute sulfuric acid leaching, and electro-

lysis(28).

Through a correspondence with Mr. S. Tsunoda, the follow-

ing principles were explained(29):

A. Elimination of ferrous sulfate in the solution may

normally be affected by converting ferrous sulfate into ferric sul-

fate and then neutralizing it to form a precipitate of ferric

hydroxide.

2FeS04 + HZSO4 +·é 02-——>'Fe2(S04)3 + H20 ................(32)

Fe2(SO4)3 + 3Ca(0H)2 -——>·2Fe(0H)3 + 3CaS04 ..............(33)

40

This process is not favored, however, because the fine precipitate

of hydroxide will make filtration difficult, and the large volume

of hydroxide precipitate will cause a considerable amount of loss

of the valuable metals by adhesion.

B. In solution with a proper amount of HZSO4, ferrous sul-

fate can be precipitated in the form of basic ferric sulfate by

addition of manganese dioxide powder. The reaction is:

6FeS04 + 3Mn02 + HQSO4 + nH20-e>3Fe203•4S03•nH20 + 3MSO4 + H20

...............(34)

C. If an excess amount of H2S04 is present in the solution,

the precipitation will not occur even though the ferrous sulfate

will be oxidized to form ferric sulfate as the following reaction:

2F€SO4 + 111102 + ZHZSÜA(35)

D. If only a slight amount of HZSO4 is present, a precipi-

tate of basic ferric sulfate may be formed:

lOFeS04 + 5MnO2 + HZSÜA + nl-120 ——>5Fe203•6S03•n.H20 + 5MnSO4 + H20

...........(36)

E. Although the mechanism of these reactions is not clearly

understood, the fact is that ferrous sulfate can be oxidized by

manganese dioxide in an acidic solution and precipitated in the

form of basic ferric sulfate, x(Fe203)•y(S03). Ferrous sulfate is

then eliminated from the pregnsnt solution. By this process, the

iron content in the pregnant solution is reduced from 5-6 grams

per liter to 1.5-2 grams per liter.

41

Applications of Fluidized-Bed Roasting and

Dilute Sulfuric Acid Leaching

Okayama Smelter, Dowa Mining Co. Ltd,, Japgg(30)

This smelter has successfully applied the combined system of

fluidized-bed roasting, dilute sulfuric acid leaching and preci—

pitation processes to produce copper, iron ore and sulfuric scid

from the complex pyrrhotite ore mined in the Yanahara mine of the

company.

The Dorrco FluoSolids roaster is used for sulfatizing roast-

ing of the raw ore, which is finely crushed and ground to a size

range of 100 to 200—mesh and prepared into a feed slurry of 70-

75% solids. The optimum range of roasting temperature is determined

as 62s•6so°c., and the optium $02 concentration of the roasting

gas is determined as 7-8%. Rated capacity of the roaster is 115

tons per day.

The calcine is leached by dilute sulfuric acid solution (2%

H2S04) in quench tanks and agitators. The pregnant solution, which

contains 2.5 grams per liter Cu, 20 grams per liter Fe and 5 grams

per liter free HZSOA, goes to cementation troughs where the copper

is precipitated on scrap iron. The leach residue is then sintered

to produce 60% iron ore, containing less than 0.2% Cu, and sent to

Amagasaki steel plant for treatment. Ninety-eight per cent sulfuric

acid is produced from the exhaust gas of the roaster by a Mbnsanto

contact system.

42

At present the approximate average copper analyses of feed,

calcine, and leach residue are 0.80%, 1.00% and 0.14% respectively.

Copper solubility is 85% and recovery is 82%. Overall iron recovery

is 99.0%.

Kosaka Smelter, Dowa Mining Co. Ltd., Japan (28)

This smelter treats a complex mixture of copper and zinc sul-

fide concentrates produced from the Hanaoka mine of the company.

The ore mined in the Hanaoka mine is an extremely complex sulfide

ore, so-called "Black 0re," containing Au, Ag, Cu, Zn, Pb, Cd, etc.,

as the essential metallic components. Earlier tests on differential

flotation of the ore indicated it would be very difficult to separate

the copper and zins sulfide minerals individually. However, appli-

cation of the hydrometallurgical process combining sulfatizing

roasting with the Dorrco Pluosolids roaster and leaching with

aqueous solution of sulfuric acid and precipitation of valuable

metals by electrolysis has made effective beneficiation of the ore

possible. A mixture of bulk sulfide flotation of Cu-Zn concentrates,

containing 8-10% Cu, 14-18% Zn, 20-25% Fe, and 35% S, is used as

the roaster feed. The roasting conditions are:

A. Height of fluidizing layer ............. 6.5 ft.

B. Roasting temperature ................... 670°C.i 5°C.

C. Ascending velocity of roasting gas...... 1.3 ft./sec.

D. Gas quantity ........................... 8000 ft.3/min.

43

E. Gas composition ....................... 802 4-5%

803 0.4-0.5%

the remainder is N2.

F. Air supply ............................ 8500 ft.3/min.

G. Air pressure .......................... 4.5 psi.

H. Pulp density of the feed .............. ' 70% solids.

The capacity of the system is 2500 tons of ore per month.

Since about 85% of the feed is minus 150-mesh, most of the over-

flow calcine is plus 150-mesh and most of the fine calcine carried

over the top of the roaster is minus 200-mesh. The calcines, in-

cluding overflow calcine and carried-over calcine, are leached by

a dilute sulfuric acid solution (5-6 grsms per liter HZSQA) in a

series of quench tanks and agitators. At a temperature above 70°C.

manganese dioxide powder is added to the agitator to oxidize dive-

lent iron and precipitate it in the form of basic ferric sulfate

so that the iron content in the pregnant solution is reduced from’

5-6 grams per liter to 1.5-2 grams per liter. Separation of the

pregnant solution and residue is affected by a series of thickeners.

The pregnant solution, containing 50 g/1 Cu, 100 g/1 Zn, 2 g/l Fe,

1 g/1 H2804, is sent to the electrolysis plant to extract copper,

zins, and cadmium step by step. The leach residue is about 45%

of the roaster feed. The residue contains 3 g/t Au, 350 g/t Ag,

1% Cu, 2% Zn, and 45% Fe. Among them, 30% of the Au, 50% of the

Ag, 40% of the Cu and 60% of the Zn are recovered by flotation

44

processes. It is reported that further utilization of the tailings,

which contain 2 g/t Au, 120 g/t Ag, 0.6% Cu, 1-1.5% Zn, 4% Pb, and

45% Fe, is under investigation.

ggilu Metallurgical Plant, Union Miniére Du Haut-ggtangg, ggtangg(31)

This is a modern hydrometallurgical plant in Katanga, Africa.

Situated near the Rolwezi mines and concentrator of Union Miniére

du Haut•Katanga, the facilities occupy an area of one square kilo-

meter. Yearly production is a little more than 100,000 tons of

copper metal and 4,000 tons of electrolytic cobalt. Plant feed is

composed of oxide concentrates mainly malachite, and sulfides, mainly

chalcocite, associated with cobalt oxide and sulfide. The average

composition of sulfide concentrates is 45% Cu, 2.5% Co while the

composition of oxide couceuttatcs is 25% Cu and 1.5% Co. The

gangue is either siliceous or dolomitic and contains impurities as

iron, nickel and zinc. Processing is by fluidized-bed roasting,

dilute sulfuric acid leaching and electrolysis. Oxide concentrates

can be dissolved directly, but insoluble sulfide concentrates are

roasted to soluble sulfates by sulfatizing roasting in the Fluo-

Solids-type reactors. Roasting temerature is maintained at 675°C.

by an automatically regulated water spray. Approximately 98% of

the copper and 96% of the cobalt is made soluble by this process.

Sulfates and oxides are leached simultaneously with spent electro-

lyte from the copper electrolysis section. Leaching occurs in four

rows of four 80·cubic meter Dorr agitators in a four·hour period.

45

After leaching, the solution contains 60 3/1 of copper and 17 g/1

of free sulfuric acid. Leaching of oxide concentrates requires the

addition of sulfuric acid, a byproduct generated during electrolysis

of copper sulfate from sulfide concentrates. Thus a proper balance

of oxide and sulfide concentrates requires only moderate addition

of fresh acid. Leached pulps are desanded and thickened. Waste

gangue is washed counter-flow with fresh water in a series of

washer thickeners before rejection as tailings. The pregnant solu-

tion is clarified just before electrolysis. During electrolysis

copper is deposited while cobalt and associated minor elements stay

in solution. Cathodes, containing 99.7% Cu, are collected after

six days, and the speut electrolyte containing 25 g/1 of copper

with 70 g/1 of sulfuric acid goes back to the leaching plant. A

part of the spent electrolyte of copper processing circuit is bled

to the eobalt processing circuit to extract cobalt.

46

III. EXPEQLEEETAL APPRDACH

Purpgses

The purposes of this investigation were concentrated on

investigating the following effects in the process of treating

the Great Gossan Lead complex sulfide ore with oxidizing and

sulfatizing roasting in a fluidized·bed kiln and dilute sulfuric

acid leaching in an autoclavez

A. The size effect of solid particles in the fluidized-

bed kiln.

B. The size effect of solid particles in the autoclave.

C. The effect of pulp density in the autoclave.

The ultimate objective of the whole series of investigations

concerning the Great Gossan Lead ore deposit is to produce commer-

cial iron ore from the complex sulfide ore, namely pyrrhotite.

Dr. G. C. Mcßride of the United States Steel Corporation has

suggested a specification for the composition of an iron ore suit-

able for blast furnace feed as follows(32).

Copper to iron ratio 1 : 1000

Zinc to iron ratio 1 : 10000

Sulfur to iron ratio 1 : 1000

Lead to iron ratio 1 : 1000 or 1 : 10000(depending on markets)

Assuing that a 58% iron ore could be produced, this specifi-

cation would permit the following amounts of impurities:

47

Copper 0.058%

Zinc 0.006%

Sulfur 0.0587.

Lead 0.058% or 0.006%

Experimental Procedure

Principal gerimental Stages

The investigation included five principal stages:

A. Preparation of ore sample.

B. Screen analysis.

C. Roasting.

D. Leaching.

E. Quantitative chemical analysis.

The flow diagram of the experimental procedures is shown in

Fig. 3.

Descriptions of__t_h_g_Principa_l Procedures and Apparatus

A. Preparation of Ore Sample. The raw ore used in this

investigation was obtained from the Betty Baker area of the Great

Gossau Lead, about 6.5 miles north of Hillsville, Virginia.

Chemical analysis of the head sample shows its composition as:

35.52% Fe, 25.37% S, 1.96% Zn, 0.63% Cu, 0.34% Pb, and 36.18% gangue

material.

The raw ore, as received for testing, was first crushed to

minus l4•mesh. A size separation was made with the Raymond pneumatic

48

classifier. The undersize (-65 mesh) from the classifier was con-

centrated by bulk sulfide flotation with a Wemco batch flotator while

the oversize (+-65 mesh) was treated by flowing-film concentration

with a Wilfley table. The bulk sulfide flotation concentrates and

the table concentrates together with middlings were then combined

to make up the fluidized·bed kiln feed.

The bulk sulfide batch flotation was conducted with the follow-

ing specifications:

1. Pulp density: 33% solids, that is, l kg. of sample plus

two liters of tap water.

2. Reagents:

a. Concentrated sulfuric acid: 3.5 ml.

b. Sodium ethyl xanthate (AC 325): 0.2 g. (Dissolve

1 gra of the reagent in 50 ml. of distilled water,

and use 10 ml. of the solution for each batch.)

c. Pine oil: a few drops.

d. Frother AC 77: a few drops.

e. Froth collection time: about 30 minutes.

By these specifications 67% of the ore was recovered as concentrate

while 33% of the ore was rejected as tailings.

B. Screen Analysis. Samples of the fluidized—bed kiln feed

and calcines, both of the overflow calcine and cyclone calcine,

were taken for screen analyses. The Tyler standard 8-inch diameter

49

ORB SLHPLE( 5-12 um )

JLU C%!SHER

I ROLL OLESHBR+14 lE8H—<—-VIBBATING SOHBBI

I-].4 IESH

HAYIOIID OLLSBIFIER

+65 EH -65 XBSH¢ izGRAVITY OOIICENTRATION BULK SULFIED PLOTLTIOH

UILFIEY TABLE UEIC0 ELTORTLI S HDD BGS CONC TBB OONC '1'ES TLIL S

( usw ) ( usw )COIZBINED il OONCENTRATESSCREEN LNALYBIS

1'I.zU'IDIZED-Bil) ROLSTIN0OVERPLOUJCALOIRBS OYCLONE IOLLCINES

SCREEN LNLLYSIS SCREEN LHLLYSISCOHBIHEDL CLLCINES

SIZIN0

-10 - +35 lll! -35 - +65 HH -65 • +150•:~; -150 [BSH

AÜ‘LLYB\LLEACÜiFILTBLTIOI

PREGNLIT SOLUTIONS lBS1kIBS( uxssmvx suumas ) ( 1•¤os1>1¤c·1·1wL um oa: )

QUAKTITLTIVE OHEKICAL LHALYSIS

HG. 3• YIHISHBIT O! EIPERIIIIITAL PBOCBDURES

50

testing sieves and the Ro-Tap testing sieve shaker were used to

perform the screen analyses.

C. Rnasting. Roasting was conducted with a specially designed

fluidized-bed kiln. The kiln was constructed of No. 316 stainless

steel, 4-inch inside diameter. Heat was supplied to the kiln by

two electric heating coils which were capable of approximately

l750°F. and rheostat controlled. The kiln was insulated with magne-

sis, which is efficient to a working temperature of 1850°F. Close

temperature monitoring was made possible by using a Leeds & Northrup

potentiometer and a chromel-alumel thermocouple inserted in a

sealed thermocouple well. The temperature of the fluidized-bed was

controlled by varying the voltage applied to the heating coils. The

kiln was also fitted with a water injection system which made it

possible to regulate the temperature of the reaction chamber within

the operating range. In order to permit the air to enter the kiln

and also support the bed, a No. 304 stainless steel plate perforated

with 1/32•inch holes on 1/4-inch centers was placed between the

reaction chamber and the air chamber. Air supply was controlled by

a needle valve. Ore was fed into the kiln through a variable speed

screw feeder. The coarser calcines were collected from the overflow

outlet while the finer calcines were collected by the cyclone. The

exhaust roasting gas was discharged through the flue. Fig. 4 illu-

strates the construction of the fluidized-bad kiln.

51

.‘·——II #(1)

I_g2)

Y Y(15)I

I _I—»AIR TO um * 2:

‘|(13)

I(4) 3

24·· 4ruißins Ä6) (14)B)Ü

gk ~ ,211)I I

4:1..1(1)ume IHJECTIOB (8) ua mm

(2) wuxa umen (9) scmaw mm(3) 0001.1:110 0011. (10) AC vamm arm uomon(4) (11) VARIALBLE mmuarommx(5) THEHIOUOUPLB IELL (12) AIR PRESSURE OAUGE(6) mmm 0011. (13) crcnomsz(7) rasssmm ups (14) am ovxavww

(15) rum

!'IG• 4• COHSTHUCTIOI OF THE FLUIDIZID-BED XIII

52

After the roasting operation was finished, the overflow cal-

cines and the cyclone calcines were mixed together. For the subse-

quent leaching tests, the mixed calcines were then sized into four

ranges: + 35-mesh, -35 to + 65-mesh, -65 to + 150-mech, and -150-

mesh.

D. Leaching. The calcines obtained from the roasting stage

were treated by dilute sulfuric acid leaching with a one-gsllon

stirred eutoclave. The autoclave was made with all wetted parts

constructed from No. 316 stainless steel. It included heater,

cooling coll, sampling tube, safety head assembly, thermocouple

well, pressure gauge, charging openings, gas openings, turbine·

type agitator, Vent Valve, sample Valve and electric drive motor.

The construction of the autoclave is shown in Fig. 5. This well-

equipped unit was capable of both high pressure and high temperature

up to e maximum of 3000 psi. and 650°F. respectively. Heat for the

autoclave was supplied by three 1000-watt electric heating elements

which were controlled with a variable transformer. The temerature

inside the autoclave was read with a Leeds & Northrup potentiometer

and a chromel-alumel thermocouple. The pressure in the vessel was

always indicated by the pressure gauge which was connected through

the oil cylinder to the autoclave. The turbine-type agitator pro-

vided with a hollow shaft was used in conjunction with the removable

bsffles in the vessel. When in operation, a low pressure area is

created at the turbine impeller. The gases are drawn down through

53

the hollow shaft and dispersed through the liquid. The bubbles

are broken up by the baffle bars. This type of agitation provides

uniform suspension and insures constant circulation of the reactant

gases through the liquid.

After leaching, the pulp was filtered with a vacuum pump and

a porcelain funnel with No. 3 Whatman filter paper. The residual

iron ore was washed into basins and dried in a steam heating room

overnight. It was then allowed to naturally air·dry for two or

three days and weighed repeatedly until a nearly constant weight

was obtained. The air-dried residue was bottled, and samples were

taken from the residue for cheical analysis. A sample of the

pregnant solution was taken and reservezi for further study.

E. Quantitative Chemical Analysis. Quantitative chemical

analyses were conducted to determine the moisture, ignition loss,

insoluble matter, sulfur, iron, copper and zinc contents of the

fluidized·bed kiln calcines and the leach residues.

The gravimetric method by precipitatiou of barium sulfate from

hot solutions was followed to determine sulfur.

Titrimetric methods were used for the determinations of iron,

copper and zinc. For determination of iron, the process of oxida-

tion with potassium dichromate was followed. The iodide method

was applied to determine copper. The determination of zinc was

made by titration with potassium ferrocyanide in acid solution with

uranium nitrate as an outside indicator.

54

PRESSUHE GLUGE Ä/HOTOR SELL LSSEMBLYP I FIXED PINH ELECTRIC MOTOR!—

iI _§IimiE_-._.II.T DRIVEOIL CYLINDER — g 1/3 HP 110 VOLT

I 5 2,I:E I sxmm pmsx" In

I.OLSKET- 316 8S•Ei" V!

BODY ·· 316 88• I"‘¤IL

‘*‘‘‘““—‘ITEEUOIELL •- 316 SS.

ä ICOOLIIG OOIL - 316 88. IIIII ZI--?

~

I'||iII·s%· I 'SLIIPLIEG TUB! ·-3168S• I ELECTRIC HEETINC

LCITLTOR EHLPT I gpüpn Ia I JACKET

I www maTURBINE 1¤1·1•1·1•ox I I Lssmmu - 316 sa.

I5”

IBLOFPIPE ·- 316 SS. I I

I II

I I

I II II II II*—l‘—*‘—*‘I·—·

-J I..

HLX. IIOKIBO PRESSURE 3000 PSI.

IAX, IOKIH0 '1'Em>ERL'1'URE 650°F•

HG. 5• CONSTBUCTION OP THE STLNDLRDONE GLLIOH STIRRED LUTOCLLVE

55

Details of the chemical analysis methods used in this experi·

ment are presented in Appendices A through F (33¤34»35¤36).

Experimental Results

Data of Screen Analysis

The data of screen analyses on the samples of the fluidized-

bed roaster feed, overflow calcine and cyclone calcine are given

in Tables 1, 2 and 3. Table 4 shows the size distribution of the

mixed cslcine, which is calculated from the data given in Tables 2,

3 and 5.

Figures 6 and 7 show the cumulative logarithmic diagrams and

the direct logarithmic diagrams of the screen analyses on roaster

feed and calcines. They are based on the data of Tables 1, 2, 3

and 4.

goastigg Performance

The fluidized·bed kiln was operated continuously for 37 hours

and 45 minutes. Feeding, air supply and heating were started at

the same time. At the beginning, it took two hours to heat the

fluidized—bed up to the proposed roasting temperature, 1450°Fahren-

heit. The kiln was then allowed to run st constant temperature for

one hour and a quarter before the collection of cslcines was started.

The proposed roasting conditions were: feeding rate, 8 gramsl

min.; air supply, 1.25 cu. ft./min.; air pressure, 2 psi.; and roast-

ing temperature, l450° i 50°F. However, the actual performance data,

56

calculated on the basis of the arithmetic means of the experimental

records taken at 15 minute intervals throughout the whole operation

time, indicated that an average feeding rate of only 6.8 grams/min.

and an average roasting temperature of l442°F. were obtained.

When the roasting was stopped and the kiln was cooled down to

room temperature, the semi-roasted material remaining in the kiln

was collected and weighed. This weight was assumed to be represent·

ing the amount of solid particles remaining and fluidizing constantly

in the fluidized bed when the kiln was in operation. Therefore the

average retention time of the solid particles in the fluidized-bed

kiln could be estimated by dividing this weight with the actual

average feeding rate. By this calculation, the average retention

time of the solid particles in the fluidized-bed kiln was estimated

to be 6.62 hours. The detailed performance data of the roasting

experiment are listed in Table 5. The distribution of overflow

and cyclone calcines obtained from the roasting stage is shown in

Table 6.

gggghing Performance

As shown in Table 7, a factorial arrangement was adopted for

the leaching stage. The factorial experiment dealt with two vari-

able factors: size of the solid particles and pulp density. Four

levels were chosen for each of the variable factors. The four

levels of solid particle size were: (A) + 35-mesh, (B) -35 to + 65-

mesh, (C) -65 to + 150-mesh, (D) - 150-mesh. Throughout this

57

experiment, pulp density was represented by the weight of solids

treated in each leaching test. The quantity and quality of leach-

ing solutions for each leaching test were kept constant. Two liters

of distilled water plus 13 milliliters of 98% concentrated sulfuric

acid were used as leaching solution for each test. The four levels

of pulp density were therefore selected as: (1) 100 grams, (2) 150

grams, (3) 200 grams and (4) 250 grams of solids.

One test for each combination of the two variable factors

was conducted. Tbtslly, sixteen leaching tests were included in

this leaching experiment.

Except the above·mentioned two variable factors and the con-

centration of dilute sulfuric acid leach solution, the following

conditions were used for all of the sixteen tests:

A. leaching temperature: 4S0°F.

B. leaching pressure: no additional pressure except the

increase of pressure in the autoclave due to the increase of tempera-

ture from room temperature to aso°r.

C. leaching time: 2 hours, single leaching.

The leaching temperature and pressure were selected on the

basis of a previous investigation record which indicated that a

higher leaching efficiency of zinc was obtained at ¢so°r. than at

250°F. while no additional pressure was applied to the autoclave

except the pressure increase due to the elevation of temperature(5¤ 37).

58

The pressure gauge of the autoclave was a 3 1/2-inch dial guage

rated 0 to 5000 psi. and the smallest pressure value per dial divi-

sion was 100 psi.. Because the pressure guage was not sensitive

enough, no record of the actual leaching pressure could be obtained

in this experiment from the gauge.

The performance data of the sixteen leaching tests are given

in Table 8.

The roasted calcines of the four size ranges used in the leach-

ing experiment and leach residues produced by the sixteen leaching

tests were chemically analyzed. The results of analyses are given

in Tables 9 and 10. The assay values of iron, copper and zinc are

given on a moisture and ignition loss-free basis. The others are

given on the natural air-dry basis.

Table 11 gives the recovery of iron which remained in the

leach residues. Tables 12 and 13 give the extraction rates of

copper and zinc which were dissolved and extracted by the leach

solutions. The total sulfur content of the leach residue is given

in Table 14. These tables are prepared in the form of the factorial

arrangement to show the effects of solid particle size and pulp

density in dilute sulfuric acid leaching.

Figures 8, 9, 10 and 11 are based on the data given in Tables

ll, 12, 13 and 14 respectively. The effects of solid particle size

and pulp density on the recovery of iron, the extraction rates of

59

copper and zinc, and the sulfur content in the residual iron ore

are clearly shown by the diagrams.

Example calculations for the recovery of iron and the extrac•

tion rates of copper and zinc are given in Appendix G•

60

Table 1. Size Distribution of the Roaster Feed.

CummlativeSize, Tyler Mesh Weight, Percent Weight, Percent

+ 14 0.1 0.1— 14 + 20 2.0 2.1

- 20 + 28 18.8 20.9

- 28 + 35 24.7 45.6

- 35 + 48 16.5 62.1

— 48 + 65 6.6 68.7

· 65 + 100 6.4 75.1

· 100 + 150 5.8 80.9— 150 + 200 7.4 88.3

· 200 11.7 100.0

61

Teble 2. Size Distribution of the Overflow Celcine.

CuuuletiveSize, Tyler Mesh Weight, Perceut Weight, Percent

+ 14 0.0 0.0

·· 14 + 20 1.2 1.2

- 20 + 28 12.4 13.6• 28 + 35 23.5 37.1

- 35 ·I· 48 19.5 56.6

- 48 + 65 12.4 69.0— 65 + 100 11.1 80.1

-100 + 150 7.5 87.6

-150 + 200 4.9 92.5

-200 7.5 100.0

62

Table 3. Size Distribution of the Cyclone Calciue.

CumulativeSize, Tyler Mesh Weight, Percent Weight, Percent

+ 14 0.4 0.4

· 14 + 20 0.7 1.1

- 20 + 20 0.5 1.6

· 28 + 35 0.4 2,0

- 35 + 48 0.3 2.3• 48 + 65 0.9 3.2

- 65 + 100 1.7 4.9

• 100 + 150 4.2 9.1

-· 150 + 200 16.8 25.9

- 200 74.1 100.9

63

Table 4. Size Dietribution of the Mixed Celcine.

(Celculated)

CumxletiveSize, Tyler Mesh Weight, Perceut Weight, Percent

+ 14 0.1 0.1— 14 + 20 1.1 1.2

- 20 + 28 11.0 12.2

~ 28 + 35 20.8 33.0

- 35 + 48 17.3 50.3

· 48 + 65 11.1 61.4·• 65 + 100 10.0 71.4

- 100 + 150 7.1 78.5

- 150 + 200 6.3 84.8• 200 15.2 100.0

64

100

so //OVERFLOW CALCINE _,

Q 1/ XJ

80 \\ / /ä//ä

/70 1E ROASTER FEED // MIXED CALCINE>• 60 / X#1_1Ki

1/3 50 1se /ä /

ä 40 1/sn ij

30 /1 CYCLONE CALCIH1/

'20 1 /1 1//10 1 x'/o

- —*" ‘"’6

8 10 14 20 28 35 48 65 100 150 200

SCREEN SCALE, TYLER MESH

1"IG. 6• CUNULATIVE LOGARITHMIC DIAGRAMS OF

SCREEN ANALYSE ON ROASTEH FEED AND CALCINE

65

40

35

Q 30

E ROLSTER FEED

E 25CALCINE•»• 20 // \_

"* · / \x crcmms [

Q / \\ cucxxm X15 [ XIXED \ /

E [ CALCINE \ /\.

[5. 1¤ ,//

5//

>< 5*5/

X///

14 20 28 35 48 65 100 150 200

SCREEI SCALE, TYLER

KESHFIG.1. DIRECT LCCARITHMIC DIAGRAMS OFSCREEN ANALYSES ON ROASTER FEED

um CALCINES

66

Table 5. !1uid:|.zed··Bed Roasting Conditions and Results.

°·=·‘=**=·=*·“·· ‘.5ää‘£‘£§ä. ..‘$‘;ä1"„i„..fFeeding Rate 8 g./min. 6.8 g./min.

Air Supply 1.25 c.f.m. 1.25 c.f.m.

Air pressure 2 psi 2.02 psi

Routing tazpersture 1450° j 50°F. l442°F.

Total operation time — 37.75 hours.

Calcine collection time - 34.50 hours.

Total feed - 15.370 kg. (100%)

Calcine recovered:

Overflow calcine· 9.110 kg. (59.27%)

Cyclone calcine - 1.188 kg. (7.73%)

Total· 10.298 kg. (67.00%)

Semi-roasted ore remelned

in the flu:I,dized··bed ki· · 2.701 kg. (17.571.)

Routing loss · 2.371 kg. (15.431)

Average retention time of

solid particles in the

kiln · 6.62 hours

*The figures listed under Actual Perforunce are calculated on thebasis of the arithuntic means of the experinmtal records taken at15 minute intervals throughout the whole operation time.

67

Table 6. Distribution of Calcine

Product Distribution, Parcent'

Overflow calciue 88.46

Cyclone calcine 11.54

Total 100.00

68

Table 7. Factorial Arrangement of the Leaching Tests.

Pulp Density, Solid Particle SizeRepresentedby Weight ofSolids. (A) (B) (C) (D)

+ 35 -35 +65 -65 +150 -150Mesh Mesh Mesh Mesh

(1) 1°° G- XAl x'B1 Xc1 Xm

(3) 200 G. XA3 XB3XE3 XD3

(‘*) 25*1 G· XA:. XB:. XC:. Xu:.

(X) denotes the observation for each combination. The8\1b8¢1'iP€8 (A); (B); (C); 8¤d (D) denote the four levels of thefactor, size of the solid particlesg and (1), (2), (3) and (4)denotes the four levels of the factor, pulp density.

These notations are called test symbols and used to denotethe specified leaching tests in all of the experimental datapresented in this thesis.

69

Table 8. Autoclave Leaching Performance.

Test* Calcine Lesch Pregnant Recovery ofSymbol Leached Residue Solution Residual Ore

g. g. ml. Z

XA1 100 89.90 1300** 89.80

XA2 150 137.25 1800 91.50

XA3 200 186.45 1310*** 93.23

XAA 250 232.40 1820 92.96

XB1 100 88.50 1800 88.50

XB2 150 137.80 1880 91.87

XB3 200 187.00 1840 93.50

XB4 250 239.65 1860 95.86

XD1 100 87.75 1930 87.75

XC2 150 133.00 1900 88.67

Xcs 200 188.90 1880 84.45

XC4 250 231.10 1820 92.44

XD1 100 73.75 1890 73.75

XD2 150 116.90 1820 77.93

ID3 200 161.80 1850 80.90

X4 250 202.70 1940 81.08

*See Table 7 for explanstion of the test symbols.

** A little leskage of Vapor from the gasket was observed.***A little leakage of Vapor from the safety head was observed.

70

*———‘ 223 2N3 3 3 2 g„22 2Q I I2 3 “ “ " " 222 E‘qRD • {J O •3 3 ¤ 3 :222 2 2¤ N ° ' ° ° 0 00 •-1 uG G Ö "" , 4,;° $**3* 3::N N N N Nägm « 2

°N ‘^

"°‘

-2 •¤ •° ° 5 5 cz M2 2 3 3 ¤ ¤ 2222 ~ 2$2*m 2-GQ x~ N >¢¤•~

¤m¤:2 3 3 3 2 2„g2 ggs5 rn N ·*‘ "" "" " 0°a$ Saw2 2222 22 = 22 ·E2uq 1x R N '• ”Qw § N Q E Q 7 2*:7 géwO

¤°‘

S ° ° °ß*«Ä°u¤+-u sus:-5 HHgnmo

SM';U!SQMQ) vll2 ··2 323•-IUH

¤•¤ 4-am vu:gg Gx Ih N W2 §~gN Q 1 1 Q SäggäääéIR N Ü" G U Q

·-• §··ö3 2-« >-.¤¤¤5·;£_¤___‘3 ***25 ~•«•·¤;:¤mg'¤ ••-¤*ä2

§ ZÜ2g2„ä~® ON <I' '•°Q,,3 *3* Q "1 °I Q •§ätn"U'g<3_:0

Ö: .°" '\ hlfn0

ggäzvähggE: ¤·*-5-¤—¤:>2¤-am2 2 E2-22°22" uÜ N 1 Q QN*3 3 3 3 22 2 355O gfäägow

N232222222 3S2¤g23gääam 4 *3* Ü ¤ ,.„•.Y«+¤00·«•¤¤^“€“§ä2V

(D2;c¤•H-aw I-•jä .-« N cn <r E*·•°°

""

71

G N N Q Q G Q an •-1 N N *13

:5 °N N •-4 N 1-I N N O •-I r-1 N N •

Q „ • • . „ „ • . • . . •

°•-• •-n r··1 «-1 •-• •—• •-·•

N ¤-1 •·—• N ""N

Q M M Q oo Q N m cn G ¤¤5 Q 'T 'T 'T **2 **2 **2 **2 *° °° 'T **2

OO O O O O O O O O O O O

""

•-4~¤ Q ND G N M N •·-1 sn M

"’*Ö

Q m os cs 61 Q cs nn 61 Q•~ Q *1 Q

I•

1 I I I I I I

F30 G12 fü N O r4_ r-6 O O O V'} QNnnen en us nn un MN Q vs nn an VN Q

o G«—•

M Q7 N cN un Q M •-1 •—•~¤

1;* „, Q 'T *2 S*€

*2 S S °) °2 °£ °*2 S °2

Ä•-1 •-1 F-• .-1 «-1 •-1 «-• G G •-1 G •"‘

*^

.¤8

«·1 G xo N M ~¤ rn Q Q M N Q NQ M M G Q G ~¤ N ***2 °:

I I I I I O I I I I

·*4-1 .5 G G GN Ge N N •-• •-4 N Q Q O •"*Q •-1 •-1 •-1 --4 •-1 •-·• H

2¤ N

3 *5U u•

E W50?) O •-I fü V1 fü fü O N ON O O "" 03fü O O N F5 v-I rz! Q füOI

I I I I I I O

7; Baß m cn cn 61 M m m 61 en 61 61 N *1

2E mQ m

0 Q cs cn Q 61 co rl Q ¤~ us co M VNou N •—l N Q UN N GN O fü O ¤-I VW

‘·n

gl• • „ • • • • • • • • • •

• fü fü N N N N N v-I fü N N N *0

c 029•-<

2Ä Uul 11H 5 N N O •—• •—l In Ih •-I N •—4 N Q M

u¤-• Q M M Q M N N G G --1 N '~¤n N

• • • • • • • • • • •° '

EG G G G G G G G G G G O Ö

•—¢UOV3„¤E-°-ää 5.* S 2.* S "‘

°' 2 ja? "‘°'

"‘ N5'ö‘

m >< :4 >< ><=·€°

>Ä°° >< Ä) >€‘° >ä’* ><

Ps

20 an Q N co oN c> •-1 61 M Q rn ~¤ *1Z «-1 •-• «-1 •-n •-• •-1 «-1

•···•

72

uN oo un 2 aj¤ <*' '^‘^

u ·•-¤N N ° ° ' an w'°

"‘ ‘°o.: vaEU A3

oo·-• M <> ETS S':}:.1 N Q

ZV"S” ”äÄ gnm M ·"‘Q HW3 3 °* S: 22U Q Qx ß

rg Q) Ü ¢ yu&W·•-•

oa ·¤*°ra ‘H "’Q ¤H• 'UO¤ ~¤2 SS*8 R ,*1 m·•-• mw:° ua N •

° ° Ü? =-••I\ ND ‘°°Qnge: Eog_•·¤ mw

1;.;X D4 H3Q (Q gg cx EIHFÄ Q."‘

~o o ¤¤ Q:°° ° N · ' ° °' arg?*;,8 ä 3

¢‘°°

.5“"«„•=Vs •·*

.:•v¤ =·'··· ä:„„ °-mE 2 :22 S:—· am *****,:*2 *62.2'° un-1mN *7*""3°•„-u

3 Q °} , ·¤>« 3;;S§·.°“

¤= M ~° E„_§ M;5 ~2„ S3m äcäß WAOusQ'; . ~= 2 2 2 643 .g2Q O·== 2 3 ~¤ ··= 2,2:2 S 38 4 ·¤ ?+-•

su G"ap-4¢A NNo ¤ <r ¤¤.¤

N __'N „„ Sun o¤•¤

• • •Q Ogg g ¤ ¤ ¤ gg; ,5;

Q 1-4>C‘•E.; ,QA 6******Naß! :>¤¤mug mmcU3 q A 0)

ä‘“S N?}

" 3 "3¤Hm *5.ugggs. ¤J3_gen IEEE;>„ S ¤¤as „¤ °¤W' Q Ü ém

QQ N ¤°¤ =

73

Table 11. The Effect of Particle Size and Pulp Density on the

Recovery of Iron in Dilute Sulfuric Acid Leaching.

Pulp Density, Solid Particle SizeRepresentedby Weight of (A) (B) (C) (D)Solide + 35 — 35 + 65 - 65 + 150 •l50

Mesh Mesh Mesh Mesh

(1) 100 G. 98.34 93.32 97.33 73.26

(2) 150 G. 99.27 96.25 97.55 78.06

(3) 200 G. 97.96 97.09 99.92 82.49

(4) 250 G. 98.50 96.05 99.66 82.08

The values of iron recovery are given in percent.

74

(A) (¤)1009°

(B) (B)

4/

E BEE 6o

{ 4Q X (B)- 35 + 65 IESK

30 Ü (C) — 65 + 150 EBK

2Q A (1)) - 150 msn

10

· 0

50 100 150 200 250

PULP DENSITY, GRAIS OF SOLIDS

FIG• 8• THE EFFECTB OF PARTICLE SIZE AND PULP DENBITYON THE RECOVERY OF IRON IN DILUTE SULFURIC ACID LEACHING

75

Table 12. The Effect of Particle Size and Pulp Density on the

Extraction Rate of Copper in Dilute Sulfuric Acid

Leaching.

Pulp Density, Solid Particle SizeRepresentedby Weight of (A) (B) (C) (D)Solids + 35 • 35 + 65 — 65 + 150 - 150

Mesh Mesh Mesh Mesh

(1) 100 G. 3.27 31.92 38.54 0.44

(2) 150 G. 1.70 26.66 30.84 6.88

(3) 200 G. 0.60 27.95 2.59 1.35

(4) 250 G. 0.60 1.31 3.66

The values of copper extraction rate are given in percent.

76

1°° • (A.) + 35 IIESH

99>< (B) -35 +651:2:611

80¤ (c) -65 +150IESH

E v¤ä A (1)) - 150 mzsu

,_ 60

äB ?°ä 49 (¤)

(B)20

10 (D)A \

Ä { E, A.0 A

50 100 150 200 250

PULP DENSITY, GRAJIB OF SOLIDS

PIC. 9• THE EFFECTS OF PARTICLE SIZE AND PULP DENSITY OlTHE EXTRACTIOH RATE OF COPPER IH DILUTE SULFURIC ACID LEACHINO

77

Table 13. The Effect of Particle Size and Pulp Density on the

Extraction Rate of Zinc in Dilute Sulfuric Acid Leachiug.

Pulp Density, Solid Particle SizeRepresentedby Weight of (A) (B) (C) (D)Solide + 35 - 35 + 65 · 65 + 150 -150

Mesh Mesh Mesh Mesh

(1) 100 G. 57.23 55.65 67.82 44.60

(2) 150 G. 53.99 51.45 66.94 28.83

(3) 200 G. 56.55 47.30 29.95 22.36

(4) 250 G. 52.47 12.51 60.98 22.93

The values of zinc extractiou rate are given in percent.

78

• (A) + 35 IIESH

100X (B)

- 35 + 65 msg

¤ (0)-65 +150lIESH90

A (0) - 150 1000080

6 7Ü Wä 60 (A>\äÄ

0 kxS 5

00ä‘° \\S6 3°

/“

20 (D)

>«10

050 100 150 200 250

0010 0000101, GRAMS 00 00110

010. 10. 000 0000000 00 PARTICLE SIZE 100 0010 0000101 00000 0100100100 0100 00 zm: IR 011000 001.00010 ACID LEACHING

79

Table 14. The Effect of Particle Size and Pulp Density on the

Sulfur Content of Lesch Residue.

Pulp Density, Solid Particle SizeRepresented (A) (B) (C) (D)by Weight of + 35 -35 + 65 - 65 + 150 • 150Solids Mesh Mesh Mesh Mesh

(1) 100 G. 1.70 1.44*),98(2)

150 G. 1.40 1.47 1.23 7.07

(3) 200 G. 1.41 1.49 0.81 6.75

(4) 250 G. 1.43 1.41 6.74

The values of sulfur content of leach residue are given inpercent.

so

• (A) + 35 HESH

IG >< (B) -35 +65 msn

¤ (6) -65 +1501msH9

A (13) - 150 usa

8+-1 (D)

\~6ä

luEä

8 4

3(A) (B)

21

(6)/ "‘0

50 100 150 200 250

_ pm 1>mws1·1·1·, emma or somws

rm. 11. um Errmcm or pmxcm sxzm um mr msxm oxum summ ooumm or man xmsxmx

81

IV. STAIISTICAL ANALYSIS OF EXPERIMBNTAL RESULTS

Dssexißüm

Since zinc is the critical impurity in iron ore, the data of

zinc extraction rates obtained from the factorial leaching experiment

are analyzed statistically to determine the significance of solid

particle size and pulp density in the leaching process.

The statistical model used in this analysis is the mixed effects

model of two•way analysis of varisce, one observatlon per combination.

The principles and test procedures given by Bowker and Lieberman in

their work, are followed to carry out this

analysis.(38)

The factor of solid particle sizes is considered as a random ef-

fect, whereas the factor of pulp densities ls considered as a fixed

effect. This implies that the experimental results will lead to con~

clusions about the distribution of solid particle sizes. But the con•

clusions about pulp densities will extend only to the four fixed lev•

els used in this experiment.

The data of the factorial experiment on zinc extraction rates,

and the analysis of variance table are given in Tables 15 and 16.

The level of significance is chosen as 0.05. From the manner

in which the model is developed, it is necessary to assume that the

interaction is nonexistent in testing for the random effects, whereas

the test for the fixed effects does not depend on any assuption

about the interaction.

82

Table 15. Data of the Factorial Experiment on Zinc Extraction Rate.

Pulp Density, Solid Particle SizeRepresented Totale Averagey Weight of + 35 · 35 + 65 · 65 + 150 • 150

Solide Mesh Mesh Mesh lhsh

(1) 100 G. 57.23 55.65 67.82 44.60 225.30 56.33

(2) 150 G. 53.99 51.45 66.94 28.83 201.21 50.30

56.55 47.30 29.95 22.36 156.16 39.04

(4) 250 G. 52.47 12.51 60.98 22.93 148.89 37.22

Totale 220.24 166.91 225.69 118.72 731.56

Average 55 .06 41 .73 56 .42 29

.68'.l‘hevalues of zinc extraction rate are given in percent-

83

Table 16. Analysis of Variance Table for Zinc Extraction Rate.

Source of Variance Sum of Degrees of Mean TestSquares Freedom Square

Between particle sizes 1900.01 3 633-34 F = 3-87

Between pulp densities 1001.18 3 333.73 F = 2.04

Residual 1471.21

_

163.47

_

Total 4372 -40 15 291 .49

-

84

jezt for the Particle Size Effects {Random Effects}

Hypothesis about the Homogeneity of Particle Size Effects

H : G2 = 0

where 62 is the variance of the particle size effects.

Rejection Area

Reject the hypothesis if,

F_ Mean Eetween gaääiclesizesquare

res ua

where the value of FO_O5(3,9) is obtained from the table of F

distribution,

F0.05(3,9)“ 3-86

.'£séL!=.F = äääfäé = 3.87$i¤°‘== F > Fo.os(a,9)

Therefore, ggjggg the hypothesis H : v2-

0

Conclusion

At a significance level of 5%, there is significant difference

in zinc extraction rate due to the difference of solid particle sizes.

85

jgs; fpr php Pulp Ep-psity Effecgg ( Fjggé §iLgg;; [

lygethesjss ppout the Hcrmocgcneitv of Pulp Densitv Effects

H¤$1=$2==$3==$4==O

where $1, $2, $3 and $4 are the effects of level (1), level (2),

level (3) and level (4) respectively.

Reject the hypothesis if,

_ gg ggggg betwgep QglgdéggitiggF“Mean square residual O‘O5(3»9)

where the value of FO_O5(3,9) is obtained from the table of F

distribution,

F0.06(a,6> = *86

Eu:.F „ .}.22.,.23.. = 2.06

163.47Since, F

4Therefore,ggggp; the hypothesis H : $1 = $2 a 53 = 54 = Q,

§ms.lu.¤.;L¤eAt a significance level of 5%, there is no significant differ-

ence in zinc extraction rate due to the difference of the four levels

of pulp density tested in this experiment.

86

V. DISCUSSION

lt is evident that the degree of roasting is directly propor-

tional to the roasting time. In fluidized·bed roasting, the roast·

ing time is the retention time of solid particles in the kiln. In

this experiment, the average retention time of solid particles in the

kiln was estimated as 6.62 hours. However, the finer particles are

easily carried through the bed by the ascending roasting gas and dis-

charged from the kiln rspidly through the cyclone or the flue. The

retention time of fine particles is shorter than the retention time

of coarse psrticles, and the fine particles may not have sufficient

roasting time to allow complete conversion of sulfides to oxides or

sulfates. Therefore overflow calcine is more desirable than cyclone

calcine.

Table 2 indicated that 87.6% of the overflow calcine was plus

1S0·mesh and 12.4% of it was minus 150-mesh; while Table 3 indicated

that 9.1% of the cyclone calcine was plus 150-mesh and 90.9% of it

was minus l50•mesh.

It is interesting to note the size distribution diagras shown

in Figures 6 and 7. From Figure 7, it is found that the overflow

calcine curve intersects the roaster feed curve at a point between

35-mesh and 48·mesh, and at another point between 150-mesh and 200-

mesh. Between these two intersections, the overflow curve lies above

the roaster feed curve. This may be explained by the overflow rate

87

of the solid particles within the size range of minus 35 to plus 150-

mesh, which is comparatively higher than the overflow rates of parti-

cles in the range of plus 35 mesh and the range of minus 150-mesh.

The vertical distance between the curve of overflow calcine and the

curve of roaster feed is largest at the point of 65-mesh, which is

the intermediate size of the range of minus 35 to plus 150 mesh.

This is identical to the fact that, in Figure 6, the cumulative size

distribution curves of overflow calcine and roaster feed intersect at

the point of 65·mesh. At this point the slope of the overflow calcine

curve surpasses the slope of the roaster feed curve. It may be ex-

plained that, aong the solid particles of the roaster feed, the

particles around 65-mesh have the highest overflowing rate,

Based on these data, it seems reasonable to determine an optimum

size range of minus 35 to plus 150—mesh, and a best size of around

65-mesh for the standard of solid particle size in the fluidized-bed

roasting process. But this inference might be effective only when

the roasting is conducted under the same roasting conditions which

were performed in this investigation.

.· §_ t 1 B .;¤ 1.·. ' ~ ·:;

3

It has been confirmed by the statistical analysis that the vari-

ance of the particle size effects on the extraction rate of zinc is

significant at a significat level of five percent. Comparing this

statistical result with the evidences revealed in Figures 8, 9, 10

and ll, it is easily found that the inference derived from the

88

statistical analysis on zinc extraction rate can be extended to the

iron recovery, copper extraction rate as well as the sulfur content

of the residual iron ore.

In Figure 8, which shows the iron recovery in the dilute sul-

furic acid leaching, the curves (A), (B) and (C) indicate that the

iron recoveries for all the plus 150-mesh particles closely approach

100 percent while the (D) curve indicates that the iron recovery for

minus 150-mesh particles lies between 73.26 and 82.08 percent. This

implies that about 20 to 30 percent of the iron content in the minus

150-mesh calcine was dissolved into the pregnant solutions. In a

beneficiation process combining leaching and electrolysis for recovery

of copper or zinc, the iron dissolved in the leach solution is a crit-

ical impurity that will affect the efficiency of copper or zinc elec-

trolysis. During the roasting process, $03 gas is formed in the

roaster. Because of the cooling action of the cyclone and the flue,

a relatively large amount of iron dust entering the hot cyclone may

be sulfated. This iron sulfate, which is collected as the cyclone

calcine, will dissolve in the dilute sulfuric acid solution during

subsequent leaching process. This general trend has been proved by

the (D) curve. It is also interesting to see the trend of (C) curve,

which shows that the highest iron recovery has been reached by the

solid particles within the size range of minus 65 to plus l50·mesh.

This experiment did not attain a favorable extraction rate of

copper. It can be found from Table 12 and Figure 9 that the copper

extraction rates of the sixteen leaching tests were lower than 40

89

percent. However, the effects of solid particle size in the copper

leaching can be revlewed clearly from Figure 9. In this chart the

(C) curve again indicates the best copper extraction rate, whereas

the (D) curve and (A) curve show the worst copper extraction rate.

It could be explalned that in the fluidised-bed roasting, the finest

particles might not have had sufficient roasting time to convert

their copper sulfide into cupric oxide or copper sulfate completely.

Previos microscopic studies on the ore indicated that the discrete

particles of chalcopyrite in the ore average less than 10 microns

and the chalcopyrite is usually finer than the marmatite. Therefore,

both in the roasting process and in the leaching process, the largest

particles might not have had sufflcient surface expoaures for complete

chemical reactions. This could be the reason that the plus 35-mesh

psrticles presented a poor copper extraction rate in Figure 9.

A consistent evidence for the highest extraction rate of zinc

for minus 65 to plus 150-mesh particles is presented by the (C) curve

in Figure 10. The (D) curve in this figure also indicates the lowest

eine extraction rate for the minus 150-mesh particles. It is evident

that the effect of surfaces becomes greater as the degree of sub-

division of matter is increased. In the fluidized bed, the zinc

oxide and the iron oxide of finer grains would have higher possibility

to contact each other and form the lnsoluble zinc ferrite at somewhat

above 650 degrees Centigrade ( equivalent to 1200 degrees Fahren-

heit ). This might be the reason that caused the lowest extraction

rate of zinc for the finest particles in the leaching process.

90

In a hydrometallurgical process with roasting and leaching, a by-

product of sulfuric acid can be produced by recovering the sulfur con-

tent of the sulfide ore. As roasting process is applied to prepare

sulfide ore, there is no difficulty to oxidize the sulfur content

with air at high temperature as completely as desired. However,

some of the sulfur is usually remaining in the calcines, both in the

forms of sulfates or sulfides due to incomplete roasting, when the

roasting is not aimed at complete oxidation of sulfur. Aa discussed

above, sulfates are easily formed in the cyclone and the flue because

of the formation of $03 gas in the roaster and the cooling action of

the cyclone and flue. The sulfate sulfur is therefore accumulated

in the cyclone calcine. Again, the cyclone calcine may contain a

lot of sulfide sulfur because it may not have had sufficient roasting

time to convert all of its sulfides into sulfate or oxides. These

phenomena can be found in Table 9, which indicates that the total

sulfur content of minus 150-mesh calcine, that is the major component

of the cyclone calcine, is 7.521 while those of the calcinea belong-

ing to the other three size ranges are 1.80 to 1.871. Sulfur con-

tent of the solid material may be reduced slightly in the dilute sul-

furic acid leaching. This tendency can be observed from the analy-

tical data of sulfur shown in Table 9 and 10. Figure 11 shows the

distribution of sulfur content in the leach residues. Iron ore is re-

quired to have a low sulfur content. In this figure, the (D) curve '

Vlies far above the (A), (B) and (C) curves. It is notable that the

91

(C) curve indicates the particles within the size range of minus 65

to plus lSO—mesh have the lowest sulfur content. This is also identi-

cal to the above-mentioned evidences of highest iron recovery and

highest copper and zinc extraction rates for the minus 65 to plus 150-

mesh particles.

From these experimental results, it can be concluded that the

solid particles within the size range of minus 65 to plus 150-mesh

would have the highest iron recovery, the highest extraction rates of

copper and zinc, and the lowest sulfur content in the leach residue.

The effects of particle size in both roasting process and leach—

ing process are closely related. The optimum size range ( minus 65

to plus 150-mesh ) for leaching process is within the optimum size

range ( minus 35 to plus 150-mesh ) for higher overflowing rate in

the fluidized-bed roasting process. The (B) curves in Figures 8,

9, 10 and ll indicate a medium preference for the particles within

the size range of minus 35 to plus 65-mesh. For practical purposes,

an optimum size range of minus 35 to plus 150-mesh could be applied

to both the fluidized-bed roasting process and the dilute sulfuric

acid leaching process.

In this experiment, an aqueous solution of sulfuric acid, con-

sisting of 2 liters of distilled water and 13 milliliters of concen-

trated sulfuric acid, was used for each leaching test. The aount of

calcine used in each test varied within the four levels, 100 grams,

92

150 grams, 200 grams and 250 grams. It had been estimated stoichio·

metrically that the concentration of sulfuric acid in the leach solu-

tion was sufficient enough to dissolve the copper and zinc contents

in 250 grams of the calcine.

The statistical analysis for the pulp density effects on the

extraction rate of ainc has concluded that, at a significance level

of 5 percent, there is no significat difference in zinc extraction

rate due to the difference of the four levels of pulp density tested

in this experiment. Becase there is a stoichiometric relationship

between the sulfuric acid concentration of the lesch solution and the

amount of acid soluble compounds in the calcine to be leached, it is

logical to consider that the inferences drswn from this experiment

can only extend to the four specified levels of pulp density used

in this experiment.

By comparing the diagrams of Figures 8, 9, 10 and ll, it can be

understood that the conclusion of the statistical analysis for the

pulp density effects on zinc extraction rate can be applied to the

pulp density effects on iron recovery, copper extraction rate and su1·

fur content in the leach residue. Almost all of the curves show a

moderate slope or even become nearly horizontal. These phenomena in-

dicate that, within the range of pulp density from 100 grams to 250

grams of solids per 2 liters of distilled water plus 13 milliliters

of concentrated sulfuric acid, there is no significant difference in

iron recovery, copper extraction rate, zinc extraction rate and sulfur

content of the leach residue due to the difference of pulp densities.

93

Hwever, some general trends of the diagrams must be noted. In

Figure 8, the general trends of the curves show that the iron recovery

decreases as the pulp density decreases. In Figure 9, the curves in-

dicate that the extraction rate of copper decreases as the pulp den-

sity increases. The curves in Figure 10 also indicate that the ex-

traction rate of zinc decreases with the increase of pulp density.

The curves in Figure ll are all nearly horizontal. lt implies that

almost no pulp density effect on the sulfur content of leach resi-

due within the specified range of pulp density.

..1;•, f 1, .5, . ‘ 4,, 9 2* .-12. g.;

Te autoclave is a pressure vessel where leaching process can

be conducted under high pressure and high temperature. But the leach-

ing process can not be carried on at an excessively high temperature

without a corresponding high pressure to overcome the vapor pressure

of leach solution. The critical point of water is 374°C. and 218

atmospheres ( about 705°F. and 3200 psi. ). As water approaches the

critical point, it becomes more like a gas and loses its solvent power

for salts.

since the autoclave was introduced to the investigation on bene-

ficiation of the Great Gossan lead complex sulfide ore at the Virginia

Polytechnic Institute, the pressure effect on the autoclave leaching

has not been studied and no record of the leaching pressure has been

reported in the previous work. Nevertheless, a previous experimental

record indicated that a higher leaching efficiency for the extraction

94

of zinc was obtained at 450°F. than at 250°F. while no additional

pressure was applied to the autoclave except the pressure increase in

the autoclave due to the elevation of temperature from room tempera-

ture to the leaching temperature. Following the previous experimental

results, the leaching experiment of this investigation was conducted

at 450°F. without applying additional pressure to the autoclave except

the pressure increase in the autoclave due to the elevation of temper-

ature from room temperature ( about 80°F. ) to the leaching tempera-

ture. Consequently, the leaching experimental results, as they were

presented in the preceding sections, did not appear to be favorable

to the extraction of zinc and the extraction of copper. The highest

zinc extraction rate attained in this experiment is 67.82 percent.

The highest copper extraction rate is only 38.54 percent. And none

of the final products, leach residues, meets the specifications of

a comercial iron ore.

A table of vapor pressure of water is given in Appendix H of

this paper. From this table, it can be found that the vapor pressure

of water at 428°F. is 336 psi~( equivalent to 22.86 atm. ). Since

the pressure gauge of the autoclave was not sensitive enough, no

record of the actual leaching pressure could be obtained in this

experiment from the gauge. Theoretically it can be inferred that

a part of the aqueous leach solution must have vaporized in the auto-

clave and lost its solvent power for salts under the leaching con-

ditions. However, leaching has also occurred at the beginning before

the temperature was raised up to 450°F., and at the end during the

95

time of cooling the autoclave. For each leaching test, it was normally

necessary to have an hour to heat up the autoclave to 450°F. at the

beginning and another one hour to cool the vessel down to room tem-

perature at the end before the vessel could be opened. But some of

the solutes dissolved into the solution during the beginning might

dissociate from the solvent when the solution vaporized. Furthermore,

the effect of the osmotic pressure of the leach solution may also

have an important role in the complicated pressure conditions of the

autoclave. The effect of leaching pressure on the experimental

results could be very important. A careful study on the pressure

effect is necessary before making a final conclusion on the appli-

cability of autoclave leaching in the beneficiation of the Great

Gossan lead complex sulfide ore.

96

IV. CONCIUSIONS

A. An optimum size range of minus 35 to plus 150-mesh and an

average size of around 65~mesh for the standard of the solid particle

size in the fluidized-bed roasting process for the complex sulfide

‘ore was determined. The solid particles within this size range have

a comparatively higher overflowing rate from the fluidized bed. But

this inference might be effective only when the roasting is conducted

under the same conditions which were performed in this investigation.

The average roasting conditions were: feeding rate, 6.8 g./min.;

air supply, 1.25 c.f.m.; air pressure, 2.02 psi.; and roasting tempera-

ture, l442°F.

B. In the dilute sulfuric acid leaching of the roasted calcine

with autoclave, the variance of the particle size effcts on the re-

covery of iron, the extraction rate of copper, the extraction rate

of zinc and the sulfur content of the leach residue is significant.

The solid particles within the size range of minus 65 to plus 150-

mesh would have the highest recovery of iron, the highest extraction

rate of copper, the highest extraction rate of zinc and the lowest

sulfur content of the leach residue. This optimum particle size range

for the leaching process is within the optimum particle size range

of minus 35 to plus l50·mesh for higher overflowing rate in the

fluidized-bed roasting process. For practical purposes, a broad

optimum particle size range of minus 35 to plus 150-mesh could be

applied to both the fluidized-bed roasting process and the dilute

sulfuric acid leaching process.

97

C. In the dilute sulfuric acid leaching of the roasted calcine

with the autoclave, within the range of pulp density from 100 to 250

grams of the calcine per 2 liters of distilled water plus 13 milli-

liters of concentrated sulfuric acid, there is no significant differ-

ence in the recovery of iron, the extraction rate of copper, the ex-

traction rate of zinc and the sulfur content of the leach residue

due to the difference of pulp densities. Hwever, within the tested

range of pulp density, the general trends are: The iron recovery

decreases as the pulp density decreases, the copper extraction rate

decreases as the pulp density increases, the zinc extraction rate de-

creases as the pulp density increases, and there is almost no pulp

density effect on the sulfur content of leack residue.

D. The pressure effect could be a significant factor in the

autoclave leaching process. It should not be disregarded. A careful

study on the pressure effect is necessary before making a final con-

clusion on the applicability of autoclave leaching to the beneficia-

tion process for the complex sulfide ore„

98

VII. SMARY

As a continuation of a series of investigations on exploring

an applicable beneficiation process for the Great Gossan lead com-

plex sulfide ore, this investigation has been conducted for the pur-

pose of studying the effect of solute size distribution on the roast-

ing and leaching of the ore. The pulp density effect in the leach—

ing process has also been studied.

The combined sulfide concentrates prepared by bulk sulfide flo-

tation and tabling were used for roasting. The oxldizing roasting

was conducted in a 4-inch diamter fluidized·bed kiln. Sulfatizing

reactions were allowed to occur together with the oxidation reactious

in the roaster. The roasting operation was carried on for 37 hours

and 45 minutes under the following average conditions: feeding rate,

6.8 grams/minute; air supply, 1.25 cu. ft./minute; air pressure 2.02

psi.; roasting temperature 1442°F.

The screen analyses on the roaster feed and calcines led to a

conclusion that, for a higher overflowing rate from the fluidized

bed, the optimum solid particle size range would be minus 35 to plus

150-mesh and the best size would be around 65-mesh. However, this

inference might be effective only when the roasting is conducted

under the same conditions performed in thh investigation.

The overflow and cyclone calcines obtained from the fluidized—

bed roasting were mixed together and sized into four size ranges for

leaching tests.

99

Sixteen leaching tests were conducted by dilute sulfuric acid

leaching with a one-gallon stirred autoclave. The leaching tests

were arranged in a factorial design which dealt with two factors,

solid particle size and pulp density. Four levels for each of the

two factors were selected. The four levels of particle size were:

(A) plus 35·mesh, (B) minus 35 to plus 65-mesh, (C) minus 65 to plus

l50·mesh, (D) minus 1S0—mesh. And the four levels of pulp density

were: (1) 100 grams (2) 150 grams, (3) 200 grams and (4) 250 grams

of solids per 2 liters of distilled water plus 13 milliliters of con-

centrated sulfuric acid. The leaching tests were conducted at 450°F.

without applying additional pressure to the autoclave except the

pressure increase due to the elevation of temperature from room

temperature to the leaching temperature.

The calcines and the leach residues were chemically analyzed to

determine their moisture, loss on ignition, sulfur, iron, copper and

zinc contents. Interpretations of the leaching results were based

on the iron recovery, copper extraction rate, zinc extraction rate

and the sulfur content of the leach residue.

Statistical analysis of the particle size effects and the pulp

density effects on the extraction rate of zinc was performed. The

inferences drawn from the statistical analysis on the zinc extraction

rate are considered to be applicable to the recovery of iron, the ex·

traction rate of copper end the sulfur content of leach residue.

From the results of the leaching tests, chemical analyses and

statistical analysis, the following conclusions have been reached:

100

A. The variance of the solid particle size effects on the iron

recovery, the extraction rates of copper and zinc, and the sulfur con-

tent of leach residue is significant. The solid particles within

the size range of minus 65 to plus l50—mesh would have the highest

iron recovery, the highest copper extraction rate, the highest zinc

extraction rate and the lowest sulfur content of leach residue.

Practically, an optimum particle size range of minus 35 to plus 150-

mesh could be applied to both the fluidized·bed roasting process and

the dilute sulfuric acid leaching process.

B. Within the range of pulp density tested in this investiga-

tion, there is no significant difference in the iron recovery, the

extraction rates of copper and zinc, and the sulfur content of leach

residue due to the difference of pulp densities. However, the general

trends are the iron recovery decreases as the pulp density decreases,

the copper extraction rate decreases as the pulp density increases,

the zinc extraction rate decreases as the pulp density increases, and

there is almost no pulp density effect on the sulfur content of leach

residue.

The purposes of this investigation, to explore the solid parti-

cle size effect on roasting and leaching and the pulp density effect

on leaching, have been attained. Bt none of the final products,

the leach residues, meets the specifications of a commercial iron

ore.

101

The pressure effect could be a significent factor in the autoclave

leachiug process. A careful study 011 the effect of leaching pressure

is reco.u.:1e11ded.

102

VIII. RECOMMENDATIONS FOR FUTURE WORK

Some modifications of the process and new topics of study are

recommnded.

S 1. .11,;,-:.: 1 9. 2, 42 ä,~=· 11

For the roasting of complex sulfide ores, oxidizing roasting at

higher temperatures will present difficulties in the prevention of

zinc ferrite formation. By sulfatizing roasting at a relatively

lower temperature near 6s0°c., it is possible to eliminate the forma-

tion of zinc ferrite. The oxides formed by oxidizing roasting require

a considerable aount of acid for dissolution in the subsequent leach-

ing process, while the sulfates formd by sulfatizing roasting are

water soluble and do nct require large quantities of sulfuric acid.

The sulfatizing roasting of complex sulfide ores with the fluidized-

bed kiln has been successfully practiced in many mines. It is pre-

ferable to oxidizing roasting.

M-! M, ;= e nased !«

For practical purposes, the open vat leaching process is simpler

and more economical than autoclave leaching. The feasibility of the

open vat leaching process for treating roasted calcines has been dem-

onstrated by its application in many mines. It is not necessary to

use the complicated and power•consuming autoclave to treat the roasted

calcines.

103

Qgyggg Pressure Legchggg ggth gtgclgyg bypassigg Roasting

Recent investigations have indicated the possibility of leach-

ing sulfide ores directly in the autoclave by introducing sufficieut

aounts of oxygen under high pressure and temperature. This process

will bypass the roasting stage normally required to prepare the sul-

fide ores for leaching. The application of the autoclave would be

justified economically if this process could be successfully de-

veloped. This approach should be investigsted.

The autoclave is a pressure leaching vessel. Whenever it is

applied, the pressure effect should not be disregarded. In this

study the effect of pressure ou the leaching results could be very

important. A careful study of the effect of pressure on the leachiug

process is necessary before a final conclusion concerning the appli-

cability of autoclaves to the beueficiation of the Great Gossau Lead

can be drawn.

1.es.b1.us.ise

Tb develop a beneficiation process for the complex sulfide ore,

it is necessary to study the method of eliminating iron dissolusion

in the leaching process. Manganese dioxide has been proven to be

applicable for this purpose at the Kosaka smelter of Japan. It could

be appllcable to the beneficiation process for the Great Gossau lead

ore.

104

The author believes that hydrometallurgy ia a promising tech-

nique for developing an applicable beneficiation process for the

Great Gossau lead sulfide ore.

105

IX. ACKNOWLEDGEMENTS

As a foreign student at the Graduate School of the Virginia

Polytechnic Institute, the author wishes to express his appreciation

to the faculty members of the university and his friends in the conr

munity of Blacksburg, Virginia, for their kindness in providing the

opportunity and environment for study during the past year.

The author owes a debt of gratitude to his course adviser,

Professor J. Richard Lucas, Head, Department of Mining Engineering,

whose guidance and encouragement made it possible to complete this

investigation.

Special acknowledgements are due to ,

instructor, who lent his knowledge and experience to this investiga-

tion, to Professor T. Carl Shelton and Professor Charles A. Beasley,

who read and corrected the manuscript of this work, and to

, who reviewed the statistical analysis of experimental

results.

Acknowledgement is also due to , Geologist, Geologi-

csl Survey of Japan, for providing helpful references concerning

this investigation.

The author appreciates the encouragement of the

, and the „ .

They helped the author settle in Blacksburg. Deep gratitude is ex-

pressed to . with whom the author lived

through the period of study at the Virginia Polytechnic Institute.

106

The author is obliged to his mother and family in his home

country• They have given him every support and encouragauent to

study abroad, and are waiting for his return.

107

X. BIBLIOGRAPHY

1. Corriveau, M. P., "Mineral Dressing Studies on the Great Gossan

Lead Ore from Carroll County, Virginia," Qelletig or V.P.].,

Engr. Experiment Station Series No. 113, V.P.I., Blacksburg, Va.

1956, 76 p.

2. Kline, M. H., and Ballard, T. J., "Investigation of the GreatGossan Lead, Carroll County, Va.," Q.S. Qgreee er Mjeee R,]. gjjg,

USBM, 1949, 39 p.

3. Ross, C. S., "Origin of the Copper Deposits of the Ducktown Typein the Southern Appalachian Regions," §.S. Qeolegreel SurveyPrgf. Pgeer 172, USGS, 1935, 165 p.

4. Coiner, R. M., and Lucas, J. R., "The Largest Undeveloped Sulfide

Ore-Body in the Eastern United States: A Beneficiation ProgressRspvrt." V¤1- VIII. N<>~ 4. V-P-I-.Blacksburg, Ve., 1961, pp. 1-6.

5. Shiuh, C. H., "Roasting and Ieaching Studies of a Complex Sulfide

0re," M.S. Thesis, V.P.I., Blacksburg, Va., 1963, 92 p.

6· ¤¤¤¤i¤. W· H-. Sir IßassPitman & Sons, Ltd., London, 1954, pp. 17-21.

7. Counselman, T. B., "Fluosolids for Roasting,"McGraw-Hill, New York, ( ).

8. Newton, J., "Pyrometa11urgy," Egtractrve Metgllurgy, Chapter 6.

John Wiley & Sons, Inc., New Ybrk 1959, pp. 275-297.

9. Ashcroft, E. A., "Sulfate Roasting of Copper Ores and Economic

Recovery of Electrolytic Copper Prem Chloride So1ut1ons,"

Trees. Am. Electroceem. Soc., Vol. 63, 1933, pp. 23-50.

10. Ashcroft, E. A., "Treating Ores or Concentrates of Copper and

Nicke1," Q.§. Per. 1,851,885, Mar. 29, 1932.

11. Peretti, E. A., "A New Method for Studying the Mechanism of1948. V¤1· 4.

p. 174.

12. Mecabe, C. L., and Morgan, J. A., "Studies on the Oxide and

Sulfate Layers Formed During Roasting of Su1fides," Jegrr_§ergle,1956, Vol. 8, p. 800A.

108

13. Thernhill, P. G., and Pidgeon, L. M., "A Microscopic Study ofSulfide Roasting," 1;gg;g_g]g§, Vol. 209, 1957, pp. 989-995.

14. Wadsworth, M. C., Leiter, K. L., Porter, W. H., and Lewis, J. R.,"Sulfating of Cuprous Sulfide and Luprous 0xide," Trggs. gxgg,Vol. 218, June 1960, pp. 519-525.

15. Lewis, J. R., Hamilton, J. H., Nixon, J. C., and Graverson, C. L.,"Oxidation of Chalcocite in Air Ccmpared with Its Oxidation inPure Oxygen,“ gg;;l;_jgghggl;, Tech. Pub., No. 2388, 1948.

16. Fursman, O. C., "Recovery of Mineral Values In Cupriferous andNickeliferous Pyrrhotite," §.S. ßgregg gg ggg; R.;. 604}, USBM,1962, 24 p.

17. Baroch, C. T., "Hydrometallurgy of Copper at the Bagdad Property,"Qgggg. gg. Elggtrggggg. §g;., Vol. 57, 1930, pp. 205-229.

18. Nii, K., Kimura, H., and Hisamatsu, Y., "Study on Zinc-Ferrite •

The Effect of Preparation Temperature of Ferric Oxide," ggg;.•. . · 1 1..1.; =.„1 ;= =. . ;. =. ...1 _· ·. neu Vol. 78No. 884, 1962, pp. 108-112.

’

19. Roggero, C. E., "High Temperature Fluid Bed Roasting of ZincConcentrates," Trggs. gg;. §gg. g[g§, Vol. 227, 1963, pp. 105-111.

20. Gupalo, Yu. P., "The Motion of Solid Bodies in a Fluidized Bed,"Inggegerng-Eiggcheskgg ggugggl, No. 2, 1962, Soviet Union,

Institute of Chemical Engineers, Vol. 2, No. 3, 1962, pp. 376-378.

21. Ge1'perin, N. I., and Ainshtein, V. G., "n the Influence of theSize and Specific Gravity of Solid Particles on Their HeatTransfer Coefficient Toward a Gas in a Fluidized Bed," gggggyg

N6- 3.1961.Institute

of Chemical Engineers, Vol. 2, No. 4, 1962, pp. 501-503.

22. Nakajima, J., Nishihara, K., and Kendo, Y., "The Application ofFluidized Bed Roasting for Hydrometallurgy · Treatment ofPyrrhotite and Hydrometallurgical Waste,"

961- 79. N¤· 897. 1963.pp. 151-156.

23. Newton, J., "Hydrometallurgy," Ex;;;cg;ve ggtgllurgy, Chapter 7,John Wiley & Sons, Inc., New York, 1959, pp. 413-448.

109

24. Umetsu, Y., Tozawa, K. and Sasaki, K., "Oxygen Pressure Leschingof Complex Sulphide Concentrate," Jour. of the Mining andMetallurgical Institute of Jappn, Vol. 76, No. 870, 1960pp. 925-929.

25. Daniels, F. and Alberty, R. A., "Properties of Dilute So1ution,"Phygical Chemistry, 2nd Ed., Chapter 8, John Wiley & Sons, NewYork, 1961, pp. 174-190.

26. Gaudin, A. M., "Interfaces Involving Solids," Flotation, 2nd Ed.,Chapter 4, McGraw-Hill Book Co., New York, 1957, pp. 86-88.

27. Daniels, F. and Alberty, R. A., "Phase Equi1ibria," PhysicalChemistpy, 2nd Ed., Chapter 10, John Wiley & Sons, New York,1961, pp. 244-248.

28. Tsunoda, S., "Hydrometallurgy of Copper-Zine Concentrateg atKosaka Smelter," Jour. of the Mining and Metallurggcal Instituteof Jappn, Vol. 72, No. 822, 1956, pp. 681-686.

29. Tsunoda, S., "Elimination of Iron by Using Mangsnese Dioxide,"Unpublished Manuscript, April 4, 1963, Dowa Mining Co. Ltd.,Tokyo, Japan.

30. Kurushima, H. and Foley, R. M., "Fluo-solids Roasting of Dowa'sYanahara Sulfides," Mining Epgineering, Society of MiningEngineers, AIME, Vol. 10, No. 10, 1958, pp. 1057-1061.

31. Saquet, J. J., Apraxine, V., Lakaye, J. and Troch, P., "KolweziMining and Metallurgical Operations of Union Miniére Du Haut-Katanga," Mining gpgineering, Society of Mining Engineers, AIME,Vol. 14, No. 12, pp. 71-81.

32. Foreman, W. E., "Great Gossan Lead Deposit from Carroll County,Va.; A Beneficiation Report,“ Unpublished Report, Va. Poly.Inst. Engr. Experiment Station, September, 1961.

33. Washington, H. S., The Chemical Analysis of Rocks, 4th Ed.,John Wiley & Sons, New York, 1930, pp. 237-242.

34. Furman, N. H., Scott's Standard Methods of Chemical Analysig,6th Ed., Vol. 1, D. Van Nostrand Co., Princeton, New Jersey,1962, pp. 529-544, 1007-1011, 1235-1237.

35. Hillebrand, W. F., Lundell, G. E. F., Bright, H. A. and Hoffman,J. I., Applied Inorganic Analysis, 2nd Ed., John Wiley & Sons,New York, 1953, pp. 248-249.

110

36. lord, N. W. and Demorest D. J., 3rd Ed.,McGraw·Hi11, New York, 1913, 192 p.

37. Coiner, R. M., Iucas, J. R. and Shiuh, C. H., "Beneficiation ofCarroll County Sulfide Ore ( Annual Progress Report ),"Unpublished, Va. Poly. Inst. kpt. of Mining Engr. and Engr.Experiment Station, September, 1962.

38. Bowker, A. H. and Lieberman, G. J., Eng;g_gg;j„gg Sgagggtics,Prentice—Ha11, Inc., New Jersey, 1959, pp. 315-332.

39. Sesrs, F. W. and Zemansky, M. W., Collggg Qyqgg, Addison-Wesley Publishing Co., Reading, Massachusetts, 1960, 3rd Ed.,P•

113

XII. APPENDICES

A. Method for Determination of oigture and Losg on Iggition

Procedure

1. Weigh about 1 gram of sample into a previously ignited

and weighed crucibln.

2. Heat in an air·drying oven at ll0°C. for 30 minutes.

3. Cool in a desiccator.

4. Weigh. The loss in weight represents the amount of

misture.

5. Ignite in a muffle furnace at lO00°C. for about 30

minutes.

6. Cool in the desiccator.

7. Weigh. The loss in weight represents the amount of loss

on ignition.

114

B. Method for Qgtermigatggn of lgsoluble ygtter

heesssä1. Weigh about 1 gra of sample into a 400-ml. beaker.

2. Dissolve it by adding 20 times its weight of concentrated

HC1 with a few drops of 5% SnCl2 solution.

3. Add 4 or 5 drops of HT.

4. Cover with a watch·glass and heat to 80° or 90°C. until

solution is complete.

5. Add a little more SnC12 solution to assist solution.

6. Filter off the residue.

7. Wash the residue 3 times with water.

8. Ignite the residue in a muffle furnace for 30 minutes.

9. Cool in a desiccator.

10. Weigh. The weight of residue represents the amount of

insoluble matter.

115

C. Meghgd fgr of §g1§gr

Regggnts

1. .. cc

a. Dissolve 80 grams of potasslum broide in just suf-

ficient water.

b. Pour 50 ml. of bromine into the saturated solution.

c. Mix well and dilute the solution to 500 ml.

2-a.Dissolve 100 gras of BaC12'2H2O in 1000 ml. of water.

zzsasésm1. Weigh about 1.374 gram of saple into a 400·m1. beaker.

2. Add 10 ml. of bromine·potassium bromide mixture.

3. Cover the beaker with a dry watch—glass.

4. let it stand 15 minutes in a cold bath with occasional

shaking of the beaker.

5. Add 15 ml. of concentrated H03 and allow the mixture

stand 15 minutes longer at room temperature.

6. Warm on an asbestos board on the steam bath until the

reaction has apparently ceased and the bromine has been

vaporized.

7. Evaporate to dryness. The cover of the beaker is raised

above the rim by means of riders.

8. Add 10 ml. of concentrated HC1.

9. Evaporate to dryness to expel the nitric acid.

116

10. Dehydrate the silica by heating in the air oven at lO0°C.

for 1 hour.

ll. Add 4 ml. of concentrated HC1.

12. 5 minutes later, add 100 ml. of hot water. Rinse the sides

of the beaker and the cover into the solution.

13. Remove the riders of the cover. Boil the sample for

5 minutes.

14. Cool for about 5 minutes.

15. Stir about 0.2 gram of aluminum powder into the solution.

Keep the beaker covered during the intervals between stirring.

16. When the iron has been reduced, the solution becoming color-

less, the sample is filtered into a 600-ml. beaker.

17. Wash the residue on the filter 9 times with hot water.

18. Make the solution acid by addition of HC1, using phenol-

phthalein indicator.

19. Add 4 ml. of HC1 ( sp. gr. 1.2 ) in excess.

20. Dilute to 400 ml. with hot water.

21. Heat to boiling.

22. Add a 10% solution of barium chloride in a fine strem from

a burette at the rate of 10 ml. in 2 to 10 minutes. The

reagent is added in slight excess of that required to react

with the sulfuric acid or sulfate. ( 10 ml. of 101 BaCl2'2Hé0

solution will preclpitate about 0.13 gram of sulfur. )

23. Place the beaker on a steam bath and allow the precipitate

to settle for about an hour.

117

24. Filter the solution through a fine grade filter paper.

25. Wash the precipitate 10 times with hot water.

26. Dry, and ignite the precipitate in a muffle furnace for

30 minutes.

27. Cool the white BaS0é precipitate in a desiccator.

28. Weigh the precipitate.

29. Calculate the amount of sulfur as followss

BaS04 x 0.1374 = gram S.

118

D. Method fo; Determination of Iron

Reggents

1. 0.2 N Standard Potassigm Dgchromsge

a. Dissolve 9.807 grams of the recrystallized dehydrated

salt of K2Cr2O7 in water and make up to 1 liter.

1 ml. of the 0.2 N standard solution is equivalent to

0.0112 gram of Fe.

2. Sggngous Chloglgg

a. Dissolve 60 grams of the crystallized salt of SnC12

in 600 ml. of concentrated HC1.

b. Make up to 1 liter with water.

3. ggosgggrig gggg · Sulggggg Acid Mlgture

a. Mix 150 ml. of concentrated HZSOA and 150 ml. of

concentrated HSPÜA.

b. Dilute to l liter with water.

a. Dissolve 60 - 70 grams of HgCl2 in 1 liter of water.

a. Dissolve 0.5 gram of the reagent in 100 ml. of water

and filter the solution.

Leeemre1. Weigh about 1 gram of sample into a 400•m1. beaker.

2. Diasolve it by adding 20 tims its weight of concentrated

HC1 with a few drops of SZ SnC12 solution.

119

3. Add 4 or 5 drops of hydrofluoric acid.

4. Cover with a watch·glass and heat to 80° or 90°C. until

solution is complete.

5. Add a little more SnCl2 solution, but do not exceed the

amount sufficient to decolorize the solution completely.

6. Filter the solution, and wash the residue on the filter 3

times with water.

7. Treat the residue as insoluble matter. ( See the method

for determination of insoluble matter in Appendix B. )

8. Add sufficient test lead to the solution to cover the bottom

of the beaker completely.

9. Cover the solution and boil it vigorously until the yellow

color has completely disappeared and the solution is colorleas

10. Cool and decat the solution into a 600-ml. beaker.

11. Wash out the remaining iron from the lead mat by several de-

cantations with water. ( Two or three 50·ml. portions of

water should be sufficient, the washings are added to the

main solution. )

12. If the solution becoms slightly colored, a few drops of

SnC12 solution are added, followed by 10 ml. of HgCl2

solution.

13. Add 15 ml. of H2S04•H3P04 mixture and 3 to 4 drops of

barium diphenylamine sulfonate indicator.

12014. Titrate with the standard potassium dichromate solution.

'Ihe green-colored solution changes sharply to a deep violetat the end point. 415. Calculate the amount of iron as followszWt. Fe in sazuple =· (Volume K2Cr207 solution in ml.) X (Fe value per ml.)

121

E. Methgd for Degegglggtgog of Cogggr

ßeasente

1-a. Dissolve 4 grams of Na2S203'5H20 in water.

b. Add 1 gram of Na0R and dilute to 1 liter.

c. Standardize the thiosulfate solution as follows:

(1) Weigh about 0.1 gra of copper foil into a 250-ml.

beaker, dissolve it in 10 ml. of HMS.

(2) Dilute to 25 ml. and boil off the red fumes.

(3) Add 5 ml. of bromine water and boil about

10 minutes.

(4) Dilute to 75 ml., and add NHAOH until the solution

becomes just alkaline but not more.

(5) Add acetic acid until the solution becomes acid,

then 2 ml. more of acetic acid.

(6) Cool to room temperature.

(7) Add 3 grams of potassium iodide dissolved in 15

ml. of water.

(8) Stir a moment.

(9) Run in the thiosulfate solution from a burette

until the yellow color of free iodine has nearly

gone.

(10) Add 5 ml. of starch solution. This should produce

a marked blue color.

122

(ll) Continue the titration until on the addition of

another drop the blue color of starch iodide

disappears.

(12) 1 ml. of the thiosulfate solution should be equal

to about 0.001 gram of copper.

2. Sggggh Sglugggg

a. Stable starch indicstor solution prepared by the

Fisher Scientific Company can be used.

a. Dissolve 125 grams of potassium iodide in water and

dllute to 500 ml.

a. Dilute 50 ml. of bromine to 500 ml. with water.

!x.9ss.@xa1. Weigh about 2 grams of sample into a 250·ml. beaker.

2. Add 7 ml. of HC1.

3. Warm.

4. Add 10 ml. of HN03.

5. Heat gently until the ore is decomposed.

6. Add 7 ml. of Hzsoqn

7. Heat until fumes of H2S0A are given off.

8. Cool.

9. Add 30 ml. of water.

10. Warm until soluble salts are dissolved.

123

11. Cool, and add 4 grams of granulated zinc.

12. Shake for 5 - 10 minutes.

13. Heat until zinc is gone.

14. Add 25 ml. of H2S water.

15. Filter, wash the residue with feebly acidulated HZS water

until iron is removed.

16. Wash the copper ( 1.e. the residue ) back into the beaker

with as little water as possible.

17. Add 7 ml. of HN03.

18. Heat to dissolve the copper, and boil to expel oxides of

nitrogen.

19. Filter the hot solution.

20. Wash the filter paper with a few ml. of bromine water.

21. Wash the beaker and filter paper with hot water.

22. Boll the filtrate until all bromine is expelled.

23. Cool, and add NRAOH until the solution just turns blue.

24. Acidify with acetic acid, and then add 3 to 4 ml. of 80%

acetic acid in excess.

25. Cool to room temperature.

26. Add 15 ml. of potasslum iodide solution, stir well.

27. lmmediately titrate with the thiosulfate standard solution.

28. When the brown tints have nearly disappeared, add 5 ml. of

starch solution, and continue the titration until one drop

changes the color from blue to yellowish-white.

124

29. Calculate the amount of copper as follows:

Wt. Cu in sample ·= (Vol. of the thio solution in ml.) X (Cu value per ml.)

125

F. Method jo; Deteggiggtiog gf Zing

Beasenß

1. ggonia Wash Solution

a. Dissolve 100 grams of NHÄC1 (C.P.) in 10001mL of

water.

b. Add 100 ml. of NHAOH ( sp. gr. 0.90 ).

2. Sulfgrgc Ac;d Q 1;; }

a. Mix 500 ml. of HZSO4 ( sp. gr. 1.84 )

with 500 ml. of water.

3. Urggggm Nitrgte

a. Dissolve 5 grams of the reagent in 100 ml. of water.

4. QL Hydrggeg Perogide Solution

a. Dilute 10 ml. of 30% H202 solution to 100 ml. with

water.

a. Dissolve 21.12 grams of the crystallized KhFe(CN)6·3H20

in water and make up 1 liter. One ml. of the solution

is equivalent to about 0.005 gram of zinc.

b. Standardize the solution as followsz

(1) Weigh 0.2 gra of zinc (C.P.) into a 600-ml. beaker.

(2) Dissolve in 50 ml. of water and 15 ml. of HC1

( sp. gr. 1.19 ).

126

(3) Add 25 grams of NHQCI.

(4) Dilute to 300 ml.

(5) Neutralized with NHAOH, using litmus paper as

an indicator.

(6) Acidify with HC1, and add 1 gram of granuated

test lead.

(7) Boil for 30 minutes.

(8) Add 10 ml. of HC1 ( sp. gr. 1.19 ) and heat to

boiling.

(9) Pour about 100 ml. into a small beaker.

(10) Add KhFe(CN)6 solution to the remaining portion

until a drop added to uranium nitrate test solution

on a spot plate gives a brown color.

(ll) Add most of zinc solution in the small beaker to

the main portion and continue addition of KaFe(CN)6

solution until the end point is just passed.

(12) Wash all remaining solution from the small beaker

into the 600-ml. beaker and finish the titration

2 drops at a time.

?.r.2.ssdm@.

1. Weigh about 2 grams of saple into a 400-ml. beaker.

2. Moisten with water.

3. Add 30 ml. of HC1 ( sp. gr. 1.19 ).

127

4. Boil gently for 10 minutes with the beaker covered.

5. Add 20 ml. of HN03 ( sp. gr. 1.42 ) and continue the boiling

for 10 minutes more.

6. Wash the cover and the sides of the beaker.

7. Add 15 ml. of H2S04 ( 1:1 ) and evaporate to dryness.

8. Cool, and add 5 ml. of HC1.

9. Add 10 grams of NHQCI and dilute to 50 ml. with water.

10. Warm until all salts are in solution.

ll. Cool, and add 1 ml. of liquid bromine.

12. Neutralize with NHAOH ( sp. gr. 0.90 ), and add 20 ml.

in excesa.

13. Heat to boiling and boil for 1 minute.

14. Filter the solution into a 600-ml. beaker.

15. Wash the precipitate 3 times with ammonia wash solution.

16. Wash the precipitate from the paper into the original beaker.

17. Add 5 ml. of HC1 and 5 ml. of 3% H202 to dissolve the manga-

nese in the precipitate.

18. Cool, and add 1 ml. of liquid bromine, neutralize with NHQOH

and add 20 ml. in excess.

19. Filter and wash as before, catching the filtrate and washings

in the 600-ml. beaker containing the first filtrate and

washings.

20. Add 1 gram of test lead to the combined filtrates.

21. Boll for 20 minutes.

22. Acidify with HC1, and boil for 10 minutes more.

128

23. Add 10 ml. of HC1, dilute to 450 ml. with water.

24. Heat to boiling and titrate with the standard solution of

KaFe(CN)6 as described under the standardization steps (9)

through (12).

Blank Degggggngtlon

1. Dissolve 25 grams of NHACI in 300 ml. of distilled water in

a 600-ml. beaker.

2. Acidify with HC1.

3. Add 1 gra of granulated test lead.

4. Boll for 30 minutes.

5. Add 10 ml. of HC1.

6. Titrate to the end point with the standard solution.

1. Calculate the factor as follows:

where A = ml. of K4Fe(CN)6 solution required for the

standard zinc solution

B = ml. of KhFe(CN)6 solution required for the blank.

W = weight of zinc in the standard zinc solution.

F = grams of zinc per ml. of KhFe(CN)6 solution.

129

2. Calculate the percentage of zinc in sample as follows:

Zzn = Q A - B [ X F X 100W

where A = ml. of K4Fe(CN)6 solution required for the sample.

B = ml. of K4Fe(CN)6 solution required for the blank.

F = grams of zinc per ml. of KAFe(CN)6 solution.

W = weight of the sample in grams.

130

G. Qgggglc Calculations for the Recovery of Fe, and the

Extggctign ßgtes of Cu and Zn

The following experimental data of the leaching test XA1 were

given by Tables 8, 9 and 10:

Weight of calcine leached = 100 g.

Weight of leach residue = 89.80 g.

Recovery of residual ore = 89.80 1

Chemical analysis of calcine A:

Mcicture and Ig. Loss = 7.79 1

Fe = 51.02 1

Cu = 0.72 1

Zn = 2.64 1

Chemical analysis of leach residue:

Moisture and Ig. Loss = 3.36 1

Fe = 53.31 1

Cu

=Zn=

131

The assay value of Fe, Cu and Zn were given on a oisture and ignition

loss-free basis. Calculate the recovery of Fe, and the extraction

rates of Cu and Zn respectively.

Solution:

89.80 x Q1 - 0.0336) x 0.5331 X wgFe !‘¢¤¤V@YY Z = 100.00 x (1 - 0.0779) x 0.5102

= 0.8980 x 1.0480 x 1.0449 x 100 *= 98.34%

89.80 x Q1 · 0.0336) x 0.0074cu exrractien rate Z == 100 -1100.00 X (1 - 0.0779) X 0.0072 X um]

= 100 - [:0.8980 x 1.0480 x 1.0278 x 100]

= 100 - 96.73 = 3.27%

Zn extraction rate 7. = 100 X 100-]

== 100 - [0.8980 x 1.0480 x 0.4545 x 100]

= 100 • 42.77 == 57.23%

132

H. Table of Absolute Veeor Pressure ef Weter *

Temperature2

Vapor Pressure Iemperature

°C. -.m. of mercur lb./in. °F.

0 4.58 0.0886 32

5 6.51 0.126 41

10 8.94 0.173 50

15 12.67 0.245 59

20 17.5 0.339 68

40 55.1 1.07 104

60 149 2.89 140

80 355 6.87 176

100 760 14.7 212

120 1490 28.8 248

140 2710 52.4 284

160 4630 89.6 320

180 7510 145 356

200 11650 225 392

220 17390 336 428

* Abstracted from §QLg§g§e§§}§;Q§ by Sears, F. W. and Zemansky, M. W.,

Addison - Wesley Publishing Co., Reading, Massachusetts, 1960,

3rd ed., p. 371.

ABSTRACT

This investigation has been carried out for the purpose of deter-

mining the effect of solid particle size distribution on the roasting

and leaching of a complex sulfide ore. The effect of pulp density in

the leaching process has also been studied.

The oxidizing roast of the combined sulfide concentrates, pre-

pared by bulk sulfide flotation and tabling, was conducted with a

fluidized-bed kiln. The calcines were leached by dilute sulfuric acid

in an autoclave. The following conclusions have been reached:

1. In the fluidized-bed roasting, the solid particles within

the size range of minus 35 to plus l50—mesh would have a higher over-

flow rate from the fluidized bed. This inference might be effective

only when the roasting is conducted under the conditions performed in

this investigation.

2. In the process combining fluidized-bed roasting and dilute

sulfuric acid leaching, the effect of solid particle size is signifi-

cant. The solid particles within the size range of minus 65 to plus

150-mash have the highest iron recovery, the highest copper and zinc

extraction rates and the lowest sulfur content of leach residue.

3. In dilute sulfuric acid leaching there is no significant

effect of pulp density within the range tested in this investigation.

The general trends are: iron recovery decreases as the pulp density

decreases, extraction rates of copper and zinc decrease as pulp den-

sity increases, and there is little pulp density effect on the sulfur

content of leach residue.